Dissertations / Theses on the topic 'Muscle Information'

During the formation of skeletal muscle, muscle cell differentiation is induced by endogenous and exogenous stimuli which cause mononucleated cells committed to myogenic lineage called myoblasts to fuse into multinucleated myotubes. The molecular mechanisms underlying this process involve myogenic regulatory factors (MRFs) which are under tight regulation. Although transcriptional regulation plays an important role, it alone can not account for the high expression levels of MRFs. In this thesis, a facet of the post-trancriptional regulation of promyogenic factors, namely MyoD and Myogenin, through the RNA-Binding Protein HuR is investigated. The presented studies show that HuR, which specifically binds to the AU-rich elements (AREs) in the 3’UTR of the aforementioned messages and stabilizes them, is cleaved during myogenesis by caspases at the site D226 which is in the HNS region. This domain normally mediates the shuttling of HuR via the import receptor Transportin 2 (Trn2); however, during myogenesis HuR is translocated to the cytoplasm where it is cleaved to generate HuR-CP1. This cleavage product then competes with HuR for binding to Trn2 which in turn prevents HuR from relocating to the nucleus and results in its cytoplasmic accumulation where it can further stabilize of the promyogenic factors and allow for muscle differentiation to proceed. Since HuR-CP1 competes with HuR for Trn2, it was interesting to examine other ligands that HuR-CP1 and HuR-CP2, the complementary cleavage product, could bind to. To this end a pull-down experiment was performed with these isoforms and extensive lists of partner proteins obtained were then compared. A biotinylation experiment was also performed with MyoD and myogenin 3’UTRs to search for RNA-Binding Proteins that could be partners of HuR or its cleavage products. Due to the time restraint, the associations from these preliminary data have not yet been validated; however an analysis of the data, shows hnRNPK, which associates with Hu proteins and antagonizes their role in neurogenesis, to bind all three isoforms as well as MyoD and myogenin 3'UTR. It would therefore be interesting to investigate some of the proteins identified in these experiments, validate the associations, and assess the roles they play, if any, during myogenesis.
Durant la formation des muscles, des stimulis endogènes et exogènes induisent la différentiation de cellules mononuclées en cellules de la lignée myogénique qui vont ensuite fusionner pour former des fibres musculaires multinuclées. Les mécanismes moléculaires entourant ce processus mettent en jeu des facteurs régulateurs de la myogenèse, les MRFs, qui sont eux-mêmes très finement contrôlés. Bien que la transcription joue un rôle très important dans ce processus, seule, elle ne peut pas maintenir le niveau élevé d'expression des MRFs. Dans ce mémoire j'étudie un aspect de la régulation post-transcriptionnelle, notamment le rôle de HuR, une protéine qui se lie spécifiquement aux éléments riches en AU (ARE) dans les régions 3’non traduites (3’UTRs) des ARN messagers comme MyoD et myogenin. Mes expériences démontrent que durant la différentiation musculaire, HuR est clivée par les caspases au niveau de l'acide aminé aspartate (D) 226 qui se situe dans un domaine important pour le mouvement de HuR du noyau au cytoplasme et l'inverse, avec l'aide de l'importine Transportin 2 (Trn2). Le déclenchement de la différentiation suscite le mouvement de HuR vers le cytoplasme, un évènement suivi par son clivage qui génère deux produits HuR CP2 et HuR-CP1. Ce dernier ayant une affinité pour Trn2, rentre en compétition avec HuR et prévient ainsi son mouvement vers le noyau. L'accumulation de HuR dans le cytoplasme permet donc de stabiliser les ARNm pro-myogéniques plus longtemps et de favoriser la différentiation. L'association de HuR-CP1 avec Trn2 m'a aussi menée à étudier les partenaires potentiels des produits de clivage par rapport à HuR. Mes expériences de Pull down ont généré de longues listes que j'ai pu comparer. Dans le but de voir si les protéines qui s'associent aux messages MyoD et myogenin se lient à HuR ou aux produits de clivage j'ai fait une expérience de biotinylation du 3’UTR des messages mentionnés. Malgrès le fait de ne pas avoir eu le temps de vérifier ces associations, j'ai pu repérer une protéine hnRNPK qui a était étudiée dans la neurogenèse et qui antagonise l'effet de protéines Hu dans ce processus. Cette protéine semble s'associer à HuR, les produits de clivages et les messages. Cela étant donné, je pense que ce serait intéressant de valider cette association en plus des autres mentionnées dans mon mémoire et d'étudier leurs importances dans la myogenèse.

During muscle contraction, electrical signals are generated by the muscle cells. The analysis of those signals is called electromyography (EMG). The EMG signal is mainly determined by physiological factors including so called central factors (central nervous system origin) and peripheral factors (muscle tissue origin). In addition, during the acquisition of EMG signals, technical factors are introduced (measurement equipment origin). The aim of this dissertation was to develop and evaluate methods to estimate physiological properties of the muscles using multichannel surface EMG (MCsEMG) signals. In order to obtain accurate physiological estimates, a method for automatic signal quality estimation was developed. The method’s performance was evaluated using visually classified signals, and the results demonstrated high classification accuracy. A method for estimation of the muscle fibre conduction velocity (MFCV) and the muscle fibre orientation (MFO) was developed. The method was evaluated with synthetic signals and demonstrated high estimation precision at low contraction levels. In order to discriminate between the estimates of MFCV and MFO belonging to single or populations of motor units (MUs), density regions of so called spatial distributions were examined. This method was applied in a study of the trapezius muscle and demonstrated spatial separation of MFCV (as well as MFO) even at high contraction levels. In addition, a method for quantification of MU synchronisation was developed. The performance on synthetic sEMG signals showed high sensitivity on MU synchronisation and robustness to changes in MFCV. The method was applied in a study of the biceps brachii muscle and the relation to force tremor during fatigue. The results showed that MU synchronisation accounted for about 40 % of the force tremor. In conclusion, new sEMG methods were developed to study muscle function and motor control in terms of muscle architecture, muscle fibre characteristics, and processes within the central nervous system.

Motor Unit (MU) innervation zones (IZs) localization is an important step in several clinical and non-clinical applications including 1) Acquisition of sEMG signal for accurate estimation of its amplitude and other parameters by avoiding placing the electrodes on IZs, 2) Accurate estimation of the EMG-Force relationship, 3) Effective injection of Botulinum Toxin in Post-stroke Spasticity near the IZs, and 4) Guiding the obstetricians to perform episiotomy during child delivery by avoiding cutting near the IZs of External Anal Sphincter (EAS) muscle. The minimal invasive way to identify the location of the IZs generally for any muscle and specifically for EAS muscle is to use multi-channel EMG signals. MU IZs can be detected from the multi-channel sEMG signals, for a fusiform muscle if the signal is acquired with an array of electrodes placed parallel to the muscle fibers, using digital signal and image processing algorithms. As most of the signal processing algorithms work on an adequate quality of the signal, thus before detecting the innervation zone it is made sure that the signal is of good quality. For this purpose, a method based on statistical thresholding of various parameters is proposed to detect the bad channels in the sEMG signals. If the number of the bad consecutive channels are more than 2 then it is suggested to acquire the signal again, otherwise each bad channel is approximated by the interpolation of its neighbor channels. As some background noise is always acquired with the EMG signal so further image enhancement techniques are used to enhance the MUAP propagation region in the spatio-temporal images and suppress the background noise. The MUAP pattern is then detected in the spatio-temporal sEMG images using multi-scale Hessian based filtering and the corresponding MU IZs are identified as the starting point of propagation of the MUAP. A software is also developed which can be used to visualize the signals acquired from EAS, detect and display the IZs and more importantly compute and display the histogram of the IZs and generate reports which will help the obstetrician while performing episiotomy during child delivery to avoid cutting vulnerable regions that may lead to fecal incontinence at later age.

Kyoto University (京都大学)
0048
新制・課程博士
博士(人間・環境学)
甲第14717号
人博第453号
新制||人||111(附属図書館)
20||人博||453(吉田南総合図書館)
UT51-2009-D429
京都大学大学院人間・環境学研究科共生人間学専攻
(主査)准教授 林 達也, 教授 森谷 敏夫, 教授 小田 伸午, 教授 田口 貞善
学位規則第4条第1項該当

Manufacturing and construction workers undertake physically strenuous activities increasing the risk of health problems, disability, and sick leave, leading to lower job attractiveness and job candidate scarcity. In the EU, up to 44 million workers are affected by workplace-related musculoskeletal disorders (MSDs), representing a total annual cost of more than €240 billion. Exoskeleton use could alleviate muscle peak loads and reduce the risks of injury of workers. This work is related to the INTERREG's project "EXSCALLERATE" which aimed to accelerate the adoption of exoskeletons among SMEs. This research presents a comparative study of using exoskeletons by workers while performing different tasks related to their job. The tests evaluate the advantages of using exoskeletons in reducing human muscle activity, thereby, reducing the fatigue and tiredness. The study uses two commercially available exoskeletons, (1) upper body exoskeleton known as Eksovest and (2) lower body exoskeleton known as LegX. For upper body, the study performed drilling tasks at shoulder height and roof drilling positions, whereas, for the lower body, virtual chair position and squatting positions are tested which involved frequent bending of knees. Besides, the experiments based on accuracies of the data collection techniques and compare three volunteer’s body muscle data acquired by EMG sensor. From these comparisons, it is found that the muscle activity can be reduced up to 60% by using these exoskeletons, hence, increasing the work life of the workforce. The results of this study will help create awareness among SMEs towards the adoption of exoskeletons.

Breast awareness is a topic related to women's daily life. The purpose of this project is to explore the behaviors of knowing breasts from the perspective of women themselves, intending to design body experiences that facilitate the awareness of breast muscles and skin movements. The thesis begins with early-stage explorations of the meanings and methodologies of breast awareness and illustrates the entire process from preliminary research to prototyping procedure. Concepts and design methods of soma design are used to explore what kinds of interactions could be outlined in the breast sensing journey. The research process and design results propose new ways of interaction design for breast awareness, exploring the application of new material properties in soma design and breast knowing approaches. As the outcome, Sensorial Pads are made out of silicone, intending to emphasize skin tactile feeling and visual interaction for breast cognition.
Att förstå bröst är ett ämne relaterat till kvinnors dagliga liv. Syftet med detta projektär att utforska beteendet med bröstigenkänning ur kvinnors perspektiv och är avsett att utforma en fysisk upplevelse som hjälper till att förstå bröstmuskel- och hudrörelser. Denna avhandling undersöker betydelsen och metoderna för bröstigenkänning från ett tidigt skede och förklarar hela processen från preliminär forskning till prototypprocedurer. Med hjälp av konceptet och designmetoden för soma design, ta reda på vilken typ av interaktion som kan beskrivas i medvetenhet om brösten. Forskningsprocessen och designresultaten syftar till att föreslå en ny interaktiv designmetod för bröstkognition och utforska tillämpningen av nya materialegenskaper i soma design och ökad kunskap om brösten. Som ett resultat är sensorkudden gjord av silikon, som är avsedd att betona hudens brösttaktilitet och visuell interaktion.

Apesar da prática da Ginástica Acrobática ser apontada por alguns estudos como um estímulo eficiente em gerar adaptações neuromusculares associadas a uma melhora no controle postural, até o presente momento nenhum estudo investigou o comportamento das oscilações posturais de ginastas da Ginástica Acrobática frente à manipulação de informações sensoriais, como por exemplo, na presença de informações táteis adicionais. Além disso, um grande número de estudos tem mostrado um aumento nas oscilações posturais em função da fadiga muscular. No entanto, não há na literatura investigações acerca dos efeitos da fadiga muscular sobre controle postural de ginastas da Ginástica Acrobática. Visto que diferentes tipos de treinamento físico, especialmente aqueles que requerem ações rápidas e habilidosas envolvendo equilíbrio e orientação corporal, induzem melhoras no controle postural, a hipótese levantada nesse estudo foi de que os ginastas da Ginástica Acrobática deveriam apresentar, após um protocolo de indução de fadiga muscular, um aumento nas oscilações posturais menos acentuados em comparação a não ginastas (i.e. um menor efeito da fadiga muscular sobre as oscilações posturais). Dessa forma, o objetivo desse estudo foi comparar, entre ginastas da Ginástica Acrobática (grupo GYN) e não ginastas (grupo CTRL), os efeitos da fadiga muscular dos flexores plantares do tornozelo sobre o controle postural, durante tarefas de postura unipodal com e sem a presença de informações táteis adicionais (i.e. com e sem contato do dedo indicador com uma superfície externa). Além disso, foi utilizada eletromiografia (EMG) de superfície de músculos do membro inferior e do tronco para avaliar mais profundamente as adaptações do sistema neuromuscular em função da instalação da fadiga. Os resultados indicaram que, para ambos os grupos GYN e CTRL, a fadiga muscular causou aumento das oscilações posturais, enquanto o toque causou diminuição das mesmas. No entanto, para o grupo CTRL (mas não para o grupo GYN) o efeito da fadiga parece ser dependente da condição de toque, visto que quando informações táteis adicionais estiveram presentes, o efeito da fadiga sobre as oscilações posturais foi atenuado. Apesar das oscilações posturais terem sido semelhantes entre os grupos GYN e CTRL, a análise dos sinais de EMG indicou que o grupo GYN utilizou menores níveis de ativação do músculo tibial anterior (ou ativações menos frequentes), assim como diferentes distribuições espectrais dos sinais de EMG provenientes do gastrocnêmio medial, bíceps femoral e do reto abdominal durante a tarefa de equilíbrio unipodal, independentemente da condição de fadiga ou de toque
Although Acrobatic Gymnastics practice has been considered effective in generating neuromuscular adaptations associated with postural control improvements, to date no study has investigated the behavior of postural sway of gymnasts of Acrobatics Gymnastics in response to the manipulation of sensory information such as the addition of tactile sensory cues. Moreover, a large number of studies have shown increased postural oscillations in response to muscle fatigue. However, no investigation has addressed the effects of muscle fatigue on the postural control of gymnasts of Acrobatics Gymnastics. As postural control is improved by different types of physical training, particularly those requiring skilled and fast actions, the hypothesis saised in this study was that gymnasts of Acrobatics Gymnastics would be less prone to fatigue-induced postural instabilities, thereby showing a smaller effect of muscle fatigue on postural sway. Therefore, the aim of this study was to compare the effects of plantar flexor muscle fatigue on the control of single leg stance between gymnasts of Acrobatics Gymnastics (GYN group) and non-gymnasts (CTRL group), during postural tasks performed with and without additional tactile information due to contact of the index finger with an external surface. Additionally, surface EMG of the lower limb and trunk muscles was used to further evaluate the fatigue-induced adaptations on the neuromuscular system. The results indicated that, for both GYN and CTRL groups, muscle fatigue caused an increase in postural sway, while the touching an external surface decreased it. However, for the CTRL group (but not for the GYN group) the effect of fatigue seems to be dependent on the touch condition, as the effect of fatigue on postural sway was attenuated in the presence of additional tactile information. Although postural sway parameters were similar between the GYN and CTRL groups, analysis of the EMG signals indicated that the GYN group used lower levels of tibialis anterior muscle activation (or less frequent activations), as well as different spectral distributions of the EMG signals from gastocnemius medialis, biceps femoris and rectus abdominis during the unipodal postural task, regardless of the fatigue or touch condition

Sensorimotor functions and an intact neural control of muscles are essential for the effective execution of movements during daily living tasks. However, despite the ability of human sensorimotor system to cope with a great diversity of internal and external demands and constraints, these mechanisms can be altered as a consequence of neurological disorders, injuries or just due to excessive effort leading to muscle fatigue. A precise assessment of both motor and sensory impairment is thus needed in order to provide useful cues to monitor the progression of the disease in pathological populations or to prevent injuries in case of workers. In particular, considering muscle fatigue, an objective assessment of its manifestation may be crucial when dealing with subjects with neuromuscular disorders for understanding how specific disease features evolve over time or for testing the efficacy of a potential therapeutic strategy. Indeed, muscle fatigue accounts for a significant portion of the disease burden in populations with neuromuscular diseases but, despite its importance, a standardized, reliable and objective method for fatigue measurement is lacking in clinical practice. The work presented in this thesis investigates a practical solution through the use of a robotic task and parameters extracted by surface electromyography signals. Moreover, a similar approach that combines robot-mediated proprioception test and muscle fatigue assessment has been developed and used in this thesis to objectively investigate the influence of muscle fatigue on position sense. Finally, the effect of posture on muscle activity, from a perspective of injuries prevention, has been examined. Data on adults and children have been collected and quantitative and objective information about muscle activity, muscle fatigue and joint sensitivity were obtained gaining useful insight both in the clinical context and in the prevention of workplace injuries. A novel method to assess muscle fatigue has been proposed together with the definition of an easy readable indicator that can help clinicians in the assessment of the patient. As for the impact of fatigue on the sensorimotor system, results obtained showed a decrease in wrist proprioceptive acuity which led also to a decline in the performance of a simple tracing task. Regarding the adoption of different muscle strategies depending on postures, results showed that muscle activity of forearm muscles was overall similar regardless from the postures.

Venous thromboembolism (VTE) is one of the most common cardiovascular diseases. KTH and its academic and industrial partners intend to develop a system to combat VTE by forcing movements of inactive muscles. An important part of this system is a unit that can sense muscular activity over time. Electromyography (EMG) is used to measure the activation potential of muscles. The goal of this thesis is to develop an EMG device that can measure bioelectric signals and convey this data to other devices. This thesis is mainly an exploration to identify the potential solution and more work is needed to develop the required system. The EMG device must be small, modular, battery powered and be able to communicate wirelessly with other devices. A functioning EMG system requires an appropriate amplification for the result to be legible and requires extensive filtering as well as detailed circuit board design to eliminate noise or interference that can affect the result.This project utilized a top down approach. An architecture of the EMG system was made and broken down into functional blocks. Each block was implemented separately and the whole solution was tested experimentally to ensure that all the specifications were fulfilled. To validate the EMG device, a series of reference images were used together with directly observing the correlation between muscle activation and its signal with an oscilloscope.The result was a fully functional EMG device that consisted of two PCB: a PCB with EMG circuitry (analog circuit) and a PCB with digital processing for communication (digital circuit). The EMG results were consistent between test subjects and could easily be correlated to muscle movement and force. The reference images indicated that it was functioning as intended. There was still 50 Hz common mode noise present in the EMG device which could have been due to its wide bandwidth and poor low frequency properties.The goals and requirements were fulfilled: a fully functional wireless, modular, small and battery driven EMG device was developed. The noise level of the EMG could have been lower and would need some further improvements. An integrated battery could be implemented to eliminate the need for users to provide a battery. An app could be developed in tandem with the EMG device, with friendly user interface, for healthcare personnel.The thesis workers strived to minimize the number of used components and power consumption. All components were RoHS certified and discarded components were collected for proper waste management. Energy consumption could have been further minimized in the digital PCB by implementing sleep mode and a watchdog timer. This thesis strived to implement as much of the 17 global sustainability goals set by the United Nations (UN). In conclusion, the main sustainability goal of this thesis was “3 – Good Health and well-being”. Other sustainability goals were “12 – Responsible consumption and production”, “13 – Climate action”, “15 – Life on land” were deemed to have been considered in this thesis.
Venös tromboemboli (VTE) är en av de vanligaste kardiovaskulära sjukdomarna. KTH och dess akademiska och industriella partner avser att utveckla ett system med uppdrag att bekämpa VTE genom att stimulera inaktiva muskler. Elektromyografi (EMG) används för att mäta musklernas aktiveringspotential. Syftet med denna avhandling är att utveckla en EMG-enhet som kan mäta bioelektriska signaler och överföra denna data till andra enheter. Ett fungerande EMG system kräver en lämplig förstärkning för att resultatet ska vara läsbart och kräver filtrering samt utförlig kretskortdesign för att eliminera brus/störningar som kan påverka resultatet negativt.Projektet använde en Top-Down strategi. En arkitektur av EMG-systemet genomfördes och sedan delades upp i funktionella block. Varje block implementerades separat och hela lösningen testades experimentellt för att säkerställa att alla specifikationer uppfylldes. För att validera EMG- enheten användes referensbilder tillsammans med att direkt observera sambandet mellan muskelaktivering och dess signal via ett oscilloskop.Resultatet var en helt funktionell EMG-enhet som bestod av två PCB: en PCB med EMG funktionalitet (analog krets) och en PCB med digital processering för kommunikation (digital krets). EMG mätningarna var konsistenta mellan testpersoner och kunde lätt korreleras med muskelrörelse och spänningskraft. Referensbilderna indikerade att den fungerade som avsedd. Det fanns fortfarande 50 Hz common mode brus i EMG-enheten, vilket kan ha orsakas av dess breda bandbredd och dåliga lågfrekvensegenskaper.Målen och kraven uppfylldes: en fullt funktionell trådlös, modulär, liten och batteridriven EMG- enhet. Brusnivån för EMG kunde ha varit lägre och skulle behöva ytterligare förbättringar. Ett integrerat batteri kunde implementeras för att eliminera användarnas behov av att tillhandahålla ett batteri. En applikation kunde ha utvecklats för EMG-enheten, med ett användarvänligt användargränssnitt, för vårdpersonal.Examensarbetarna strävade efter att minimera användning av komponenter och strömförbrukning under arbetsprocessen. Alla komponenter var RoHS-certifierade och kasserade komponenter insamlades för korrekt avfallshantering. Energiförbrukning kunde ha minimerats ytterligare i det digitala kretskortet genom att implementera sleep mode och en watchdog timer. I detta examensarbete var det önskvärt att implemnetera de 17 globala hållbarhetsmålen uppsatta av FN (Förenta Nationerna). Sammanfattningsvis uppfylldes huvudsakligen “3 – Good Health and well-being”. Hållbarhetsmålen ”12 - Ansvarig konsumtion och produktion”, ”13 – Klimatåtgärder”, ”15 - Liv på land” anses även att ha beaktas i denna avhandling.

Le sujet de cette thèse se situe dans le cadre général de la restauration du mouvement de membres paralysés à travers la stimulation électrique fonctionnelle (FES) implantée. L'objectif du projet était d'explorer la faisabilité d'utiliser les informations issues des fibres nerveuses sensorielles des récepteurs musculaires comme information de retour d'une commande en boucle fermée d'un système FES à travers des électrodes nerveuses périphériques intra fasciculaires. Des expérimentations animales aigues ont été réalisées pour mesurer les réponses afférentes des fuseaux neuromusculaires à des étirements passifs du muscle. Les enregistrements ont été réalisés en utilisant une nouvelle électrode Intra-fasciculaire (tfLIFE), développées par le Dr. Ken Yoshida à l'université d'Aalborg au Danemark. Un modèle du premier ordre de la réponse des fuseaux neuromusculaires à des étirements passifs a été proposé. Ce modèle prend en compte les propriétés non linéaires des activités neurales afférentes. De plus, l'estimation de l'état du muscle à partir d'un enregistrement ENG multicanaux a fourni des résultats plus robustes comparés à un enregistrement monocanal.
Pour que le modèle ci-dessus puisse être utilisé pour l'estimation de l'état du muscle, le taux de variation de la longueur du muscle pendant le mouvement doit avoir un effet négligeable sur les paramètres du modèle. Nous avons proposé dans cette thèse une approche pour la détection et la classification de pics dans l'enregisrement neural dans l'objectif d'isoler les activités neurales sensorielles des récepteurs musculaires ayant une sensibilité minimale à la vitesse de l'élongation musculaire. L'algorithme est basé sur la transformée en ondelettes continue multi-échelle utilisant des ondelettes complexes. Le système de détection utilise une simple détection par seuillage, couramment utilisée, particulièrement avec les enregistrements ayant un faible rapport signal sur bruit. Les résultats de classification des unités montrent que la classification développée est capable d'isoler l'activité ayant une relation linéaire avec la longueur du muscle. Ceci constitue une étape vers une estimation, en ligne basée modèle, de la longueur du muscle qui pourra être utilisée dans un système FES en boucle fermée utilisant des informations sensorielles naturelles.
Un des principaux problèmes limitant l'interprétation des données ENG est le faible niveau du signal neural par rapport à celui du bruit dans l'enregistrement. Nos hypothèses ont été que le blindage de l'implant aiderait à améliorer le rapport signal sur bruit. Des résultats expérimentaux, issus d'une étude préliminaire que nous avons réalisée, montrent que le placement d'électrodes standards à manchon placées autour du site d'implantation de la tfLIFE augmentait le niveau du signal ENG dans les enregistrements.

Muscle force is regulated by varying two main motor unit properties: the recruitment and the firing rates of motor units. Discrepancies still exist on the mechanisms involved in motor unit control and muscle force generation. This study investigated the behavior of motor unit firing rate during sustained fatiguing contractions and the motor unit parameters that are most likely to influence force fluctuation increase. We also studied the firing rate of motor units during linearly increasing force contractions up to maximum, or near maximum voluntary contraction force, at different rates of force increase, and developed an equation that models the firing rate behavior as a function of increasing excitation to the motor unit pool. Results were used to create a model of muscle force production that is based on verifiable physiological concepts and data. The model also includes the concept of common drive, i.e. of an oscillatory common input received by all motor units in the motor unit pool, the time-dependent changes of motor unit twitches, and a feedback loop to simulate force generation in a target-force tracking mode. Simulations showed that the model is able to mimic the force and firing rate patterns which have been experimentally observed during repeated contractions sustained to exhaustion: the excitation to the motoneuron pool must be adjusted in response to an increased or decreased force generation capacity of the muscle fibers, and the firing rates of all motor units respond consequently with a decreased or increased firing rate. The simulation of prolonged contractions showed that the increase in force variability may be attributed to the gradual recruitment of higherrecruitment threshold larger-amplitude force twitch motor units. The level of cross-correlation between firing rates appeared to influence force variability, whereas the variability in the firing rates had no clear effect on force variability.
Il controllo della forza muscolare si basa principalmente su due fenomeni: il reclutamento di unità motorie e la regolazione della loro frequenza di scarica. Molti aspetti riguardanti i meccanismi coinvolti nel controllo delle unità motorie e nella generazione di forza muscolare restano ancora da investigare. Parte del lavoro di questa tesi ha riguardato lo studio del comportamento della frequenza di scarica delle unità motorie e dei parametri alla base dell’incremento delle fluttuazioni dell’output di forza durante l’esecuzione di contrazioni muscolari sostenute fino all’affaticamento. Inoltre, è stato analizzato il comportamento della frequenza di scarica delle unità motorie durante lo svolgimento di contrazioni muscolari a livelli di forza crescente fino alla massima forza di contrazione volontaria (a diverse velocità di incremento della forza); ed è stata messa a punto una equazione in grado di modellare il comportamento della frequenza di scarica in funzione dell’eccitazione ricevuta dal pool di unità motorie. I risultati di questa prima analisi sono serviti per creare un modello di produzione della forza muscolare basato su dati fisiologici verificabili. Il modello include il concetto di “common drive”, ovvero di un input oscillatorio comune ricevuto da tutte le unità motorie del pool; la dipendenza temporale dei “twitch” di forza delle unità motorie; ed un “feedback loop” per simulare la generazione di forza in contrazioni in “target-force tracking mode”. Si è dimostrato come il modello sviluppato sia in grado di simulare il pattern di forza e il comportamento delle unità motorie sperimentalmente osservati durante l’esecuzione di contrazioni prolungate e sostenute fino all’affaticamento. In particolare, si è potuto osservare come l’eccitazione ricevuta dal pool di unità motorie si modifichi in seguito ad un aumento o ad una diminuzione della capacità di produrre forza delle fibre muscolari e come la variazione dell’eccitazione comporti di conseguenza una diminuzione o un aumento della frequenza di scarica delle unità motorie e del numero di unità motorie attive. La simulazione di contrazioni muscolari prolungate ha anche evidenziato come la crescente variabilità della forza muscolare sia da attribuire al reclutamento di unità motorie caratterizzate da “twitch” di ampiezza maggiore e da un maggiore grado di cross-correlazione tra la frequenza di scarica delle unità motorie attive, mentre la variabilità della frequenza di scarica non sembra influire sull’output di forza.

In the last ten years the techniques for non invasive investigation of the neuromuscular system made consider¬able progress. Recently they have been associated to multichannel surface electromyography (EMG) detected with linear or two-dimensional elec¬trode arrays. This approach consists in measuring electrical muscle activ¬ity, generated by the most superficial motor units (MUs), with multiple (more than two) closely spaced elec¬trodes overlying either a restricted area of the skin or the entire surface of the muscle. Although the behavior of the MUs is classically studied with intramuscular EMG, these recently developed techniques allow the analy¬sis of EMG recorded in multiple loca¬tions over the skin surface. The analy¬sis of MUs from the surface EMG is useful when the insertion of needles is not desirable or not possi¬ble. Moreover, multichannel surface EMG allows the measure of MU properties which are difficult to assess with inva¬sive technology (e.g., muscle fiber conduction velocity, location of inner¬vation zones, global and single motor unit manifestations of fatigue) and increases the number of detectable motor units with respect to selective intramuscular recordings. The application of surface EMG to the study of electrically elicited con¬tractions allows a study of the com¬pound muscle action potentials (CMAPs) or massed action potentials (M-waves) which represent the syn¬chronous summation of the MUAPs that add asynchronously during a vol¬untary contraction. M-wave analysis represents an important investigative tool in several different areas of neuro¬physiological research. Evoking the maximum M-wave by supramaxi¬mal stimulation of either a peripheral nerve or the muscle motor point is the electrical equivalent of the recruitment of all motor units within the selected motoneuron pool and allows to inves¬tigate the properties of these motor units. Moreover, progressive incre¬ments of stimulation intensity elicit progressively greater M-waves. The difference between M-waves elicited using two different stimulation levels allows the analysis of the MUs whose activation thresholds are in between the two stimulation levels. This proc¬ess allows the characterization of small groups of MUs and therefore the order of recruitment and MU charac¬terization during electrical stimula¬tion. The use of incremental M-wave technique together with bi-dimen-sional detection systems is an inter¬esting application to study the spatial localization of MU territories within pinnate muscles. Electrical stimulation (combined with surface EMG) provides also in¬teresting experimental paradigms to study muscular properties during fa¬tiguing contractions, because it gives to the experimenter the control of MU firing frequency and recruitment and, if selectively applied, eliminates the problem of cross-talk from nearby muscles. In fact, selective electrical stimulation of a nerve branch or of the motor point of a muscle allows, in first approximation, to “disconnect” the investigated muscle from the central nervous system and to activate only one (or a portion of one) muscle at a time at a controlled frequency and with a motor unit pool that is more likely to be stable. During fatiguing muscle contractions, M-waves (de¬tected with linear or two-dimensional electrode arrays) change shape be¬cause of the presence of motor units with different conduction velocities and fatigue profiles. With respect to the analysis of myoelectric fatigue during voluntary contractions, electri¬cally elicited contractions provide es¬timates of the changes in surface EMG signal variables that have lower esti¬mation errors. The issue of fatigue during electrically elicited contrac¬tions is of paramount importance in functional electrical stimulation tech¬niques for external control of para¬lyzed extremities. Furthermore, the application of bursts of electrical pulses to a periph¬eral nerve (or to the muscle motor point) may induce, in certain muscles, involuntary muscle activity that per¬sists for some time after the interrup¬tion of the stimulation. Such activity is referred to as “fascicula¬tion” and “cramp”, which are defined, respectively, as “the random, sponta¬neous twitching of a group of muscle fibers belonging to a single motor unit” and “an involuntary, painful muscle contraction associated with electrical activity”. Even if cramps are a common complaint encountered by both neurologists and primary care physicians, their pathophysiology still remains poorly understood, mainly due to the unpredictable occurrence and the relative inaccessibility to experi¬mental investigation. However, sur¬face EMG signals detected during cramps induced by means of electrical stimulation of the muscle, in con¬trolled conditions of fibre length and orientation, provide interesting infor¬mation about cramp pathophysiology. This PhD thesis contributes to this research field through the analysis and interpretation of multichannel surface EMG signals during electrically elic¬ited contractions and during different types of involuntary muscle contrac¬tions, such as fasciculations and cramps.The main objectives of the thesis were to: (1) establish the influence of selected parameters of the electrical stimu¬lation (current amplitude, pulse shape, injected charge) on the de¬gree of MU activation during transcutaneous stimulation; (2) assess the differences in the myoelectric fatigue profiles during electrical stimulation among differ¬ent muscles and subject popula¬tions; (3) evaluate whether the electrical stimulation of the muscle motor point can trigger a cramp in differ¬ent foot and leg muscles, as it was already observed in foot muscles following the electrical stimulation of the nerve trunk; (4) study surface EMG signals during cramp contractions; (5) examine whether different bursts of electrical stimulation, applied to the muscle motor point, trigger dif¬ferent cramps of the muscle under study; (6) provide novel insights into cramp pathophysiology from detection and analysis of surface and intra¬muscular EMG signals during electrically elicited cramps. Part 1. NMES: Methodological issues This section focuses on methodological aspects concerning neuromuscular electrical stimulation (NMES). Chapters I and II include a brief introduction to the stimulation techniques, to the spinal involvement in electrically elicited muscle contractions, and to the acquisition of EMG signals during electrically elicited contractions (study of M-waves and incremental M-waves). Further, Chapter II reports a study on the incremental M waves detected from the medial gastrocnemius muscle. It has been preliminarly demonstrated that the use of incremental M-waves acquired with high density detection systems allows the spatial localization of MU territories in a pinnate muscle. This could have an important application in the study of the compartmentalization of muscles involved in posture, such as the gastrocnemii. Chapter III describes an experimental study on the effect of the stimulation parameters (stimulation waveform, amplitude, and injected charge) on the degree of MU activation in electrically-elicited contractions of the biceps brachii. The main outcome of this work was that the degree of MU activation was a function of the injected charge and not of the stimulation waveform. Moreover, MUs tended to be recruited in order of increasing conduction velocity with increasing charge of transcutaneous stimulation (regardless of the pulse waveform). This study concerned the biceps brachii, a muscle which is not usually electrically stimulated other than for M-wave studies. Other muscles, particularly the thigh muscles, are either of greater interest in sport sciences and rehabilitation medicine. Whether MUs are recruited in an orderly or random manner during motor point electrical stimulation of these muscles was one of the topics of the study reported in Chapter IV. The objectives of the study described in Chapter IV were to investigate, in three muscles of the thigh, the MU recruitment order in stimulated contractions and the electrically-induced manifestations of muscle fatigue. This study, performed on two groups of healthy subjects, was also aimed to test the possibility of detecting differences among muscles and subject groups in the response to the transcutaneous stimulation. Results of this study confirmed the observations of the previous study (Chapter III) on the recruitment order: MUs tended to be activated from low to high conduction velocities with increasing current of the transcutaneous stimulation. Moreover, EMG variables during electrically-evoked fatigue showed different properties in different muscles and subject groups. Overall, the results of the above-described methodological studies contributed to the design of clinical protocols for neuromuscular function assessment (sarcolemmal excitability and fatigability) in clinical relevant conditions. These studies are described in the second part of this thesis. Part 2. NMES: clinical applications Chapter V describes two experimental protocols aimed to investigate the neuromuscular effects of both glucocorticoid administration in healthy subjects and chronic endogenous hypercortisolism in pathological subjects (affected by Cushing’s disease). Results from the protocol performed on healthy subjects showed that muscle fiber conduction velocity and myoelectric manifestations of fatigue significantly decreased after the short-term administration of a synthetic glucocorticoid (dexamethasone), in doses well within the range used clinically. Results from the protocol conducted on pathological subjects showed that muscle fiber conduction slowing is a sensitive marker of steroid myopathy which is suitable to be used in combination with standard electrodiagnostic tests for identifying early signs of myopathy. Part 3. NMES induced muscle cramps The third section is focused on the study of involuntary muscle contractions. Chapter VI reports the description of a method, developed within this project, for inducing cramps in an intrinsic foot muscle, the abductor hallucis, by means of electrical stimulation of the main muscle motor point. This method was reliable and suitable to be used concomitantly with multichannel surface EMG. The test of this method in different foot and leg muscles is the topic of the second part of Chapter VI. We found that the neurostimulation method was effective in eliciting cramps in four muscles of the lower limb bilaterally, and that differences exist between the cramp elicitability profiles of the different foot and leg muscles. The analysis of surface EMG signals during cramp contractions as well as the study of the effect of different stimulation frequencies on the elicited muscle cramp is reported in Chapter VII. The aim of this study was to provide new insight into cramp pathophysiology (generation and self-sustaining mechanisms) through the analysis of EMG signals. Results showed that the choice of the frequency of the stimulation burst affects the temporal and spectral properties of EMG in electrically-elicited cramps. However, these findings did not provide insight into the exact physiological mechanism(s) underlying cramp generation since they could fit both the hypothesis of peripheral origin and that of spinal origin of cramps. Hence, the differences between cramps elicited with different stimulation frequencies were further investigated by means of a joint analysis of surface and intramuscular EMG signals (Chapter VIII). The analysis of the discharge behaviour of single MUs during cramp (discharge rate and variability, and coherence between discharge rate oscillations of different MUs) indicated that cramp development and self-sustaining mechanisms involve spinal pathways, although the origin may be peripheral. Future studies will be aimed to confirm recent preliminary evidences supporting the hypothesis of cramps being initialized at the peripheral level but subsequently supported by spinal reflex loops. HDsEMG and decomposition of cramp signals into the constituent trains of motor unit action potential represent a promising approach to investigate the behavior of individual MUs during cramp contractions.

Une définition très générale de la structure musicale consiste à considérer tout ce qui distingue la musique d'un bruit aléatoire comme faisant partie de sa structure. Dans cette thèse, nous nous intéressons à l'aspect macroscopique de cette structure, en particulier la décomposition de passages musicaux en unités autonomes (typiquement, des sections) et à leur caractérisation en termes de groupements d'entités élémentaires conjointement compressibles. Un postulat de ce travail est d'établir un lien entre l'inférence de structure musicale et les concepts de complexité et d'entropie issus de la théorie de l'information. Nous travaillons ainsi à partir de l'hypothèse que les segments structurels peuvent être inférés par des schémas de compression de données. Dans une première partie, nous considérons les grammaires à dérivation unique (GDU), conçues à l'origine pour la découverte de structures répétitives dans les séquences biologiques (Gallé, 2011), dont nous explorons l'utilisation pour modéliser les séquences musicales. Cette approche permet de compresser les séquences en s'appuyant sur leurs statistiques d'apparition, leur organisation hiérarchique étant modélisée sous forme arborescente. Nous développons plusieurs adaptations de cette méthode pour modéliser des répétitions inexactes et nous présentons l'étude de plusieurs critères visant à régulariser les solutions obtenues. La seconde partie de cette thèse développe et explore une approche novatrice d'inférence de structure musicale basée sur l'optimisation d'un critère de compression tensorielle. Celui-ci vise à compresser l'information musicale sur plusieurs échelles simultanément en exploitant les relations de similarité, les progressions logiques et les systèmes d'analogie présents dans les segments musicaux. La méthode proposée est introduite d'un point de vue formel, puis présentée comme un schéma de compression s'appuyant sur une extension multi-échelle du modèle Système & Contraste (Bimbot et al., 2012) à des patrons tensoriels hypercubiques. Nous généralisons de surcroît l'approche à d'autres patrons tensoriels, irréguliers, afin de rendre compte de la grande variété d'organisations structurelles des segments musicaux. Les méthodes étudiées dans cette thèse sont expérimentées sur une tâche de segmentation structurelle de données symboliques correspondant à des séquences d'accords issues de morceaux de musique pop (RWC-Pop). Les méthodes sont évaluées et comparées sur plusieurs types de séquences d'accords, et les résultats établissent l'attractivité des approches par critère de complexité pour l'analyse de structure et la recherche d'informations musicales, les meilleures variantes fournissant des performances de l'ordre de 70% de F-mesure
A very broad definition of music structure is to consider what distinguishes music from random noise as part of its structure. In this thesis, we take interest in the macroscopic aspects of music structure, especially the decomposition of musical pieces into autonomous segments (typically, sections) and their characterisation as the result of the grouping process of jointly compressible units. An important assumption of this work is to establish a link between the inference of music structure and information theory concepts such as complexity and entropy. We thus build upon the hypothesis that structural segments can be inferred through compression schemes. In a first part of this work, we study Straight-Line Grammars (SLGs), a family of formal grammars originally used for structure discovery in biological sequences (Gallé, 2011), and we explore their use for the modelisation of musical sequences. The SLG approach enables the compression of sequences, depending on their occurrence frequencies, resulting in a tree-based modelisation of their hierarchical organisation. We develop several adaptations of this method for the modelisation of approximate repetitions and we develop several regularity criteria aimed at improving the efficiency of the method. The second part of this thesis develops and explores a novel approach for the inference of music structure, based on the optimisation of a tensorial compression criterion. This approach aims to compress the musical information on several simultaneous time-scales by exploiting the similarity relations, the logical progressions and the analogy systems which are embedded in musical segments. The proposed method is first introduced from a formal point of view, then presented as a compression scheme rooted in a multi-scale extension of the System & Contrast model (Bimbot et al., 2012) to hypercubic tensorial patterns. Furthermore, we generalise the approach to other, irregular, tensorial patterns, in order to account for the great variety of structural organisations observed in musical segments. The methods presented in this thesis are tested on a structural segmentation task using symbolic data, chords sequences from pop music (RWC-Pop). The methods are evaluated and compared on several sets of chord sequences, and the results establish an experimental advantage for the approaches based on a complexity criterion for the analysis of structure in music information retrieval, with the best variants offering F-measure scores around 70%. To conclude this work, we recapitulate its main contributions and we discuss possible extensions of the studied paradigms, through their application to other musical dimensions, the inclusion of musicological knowledge, and their possible use on audio data

This dissertation represents the development of a novel exoskeleton (Exo-Muscle) and corresponding evaluation and controller strategies to achieve more ergonomics and intuitive operation. At the first stage of the study, we introduce the principle, design, and mechatronics of Exo-Muscle, a novel assistive device for the knee joint. Different from the existing systems based on rigid exoskeleton structures or soft-tendon-driven approaches, the proposed device leverages a new semi-rigid principle that explores the benefits of both rigid and soft systems. Using a novel semi-rigid chain mechanism around the knee joint eliminates the misalignment between the device and the knee joint center of rotation. At the same time, it forms a well-defined route for the tendon. This results in more deterministic load compensation functionality compared to fully soft systems. The proposed device can provide up to $38Nm$ assistive torque to the knee joint. The control of the Exo-Muscle relies on the knee effort computation, which is based on the forward kinematic computation with ground reaction force measurement. To achieve the multi-task operation objective, a lightweight force solely targeting long-term operation in the scenarios of the production lines is developed. The proposed design includes six sensing elements that provide a stable support polygon between the foot and the ground. To override the latent nonlinear properties of the device under high load. The Gaussian Process Regression (GPR) based non-model calibration method is proposed. In addition, a novel fault detection method is proposed based on the Gaussian distribution from a GPR model trained for fault detection to improve reliability. To bring back the device after the partial fault of the device. A recovery approach which is also based on the GPR, is proposed. The proposed recovery method can significantly improve the accuracy of extracting the center of pressure and the nominal ground reaction force from the raw sensing reading under the fault conditions. From the control perspective, we proposed a control strategy targeting improving the ergonomics and efficiency of long-term operations. To enable the long-term realistic human state evaluation, a novel EMG-free muscle fatigue evaluation method for human-assistive robot interaction scenarios. Unlike conventional methods, which rely on complex modeling of human biological joints or noisy EMG measurements, we use the GPR to reveal the latent relation between human muscle activation (via EMG measurement) and the corresponding joint effort and state during a simple calibration stage. The trained model then online approximates the muscle activation using the human joint state and effort derived from the wearable device kinematic and sensor measurements. The approximated muscle activation is used for assessing the muscle state of the subject with a model-based fatigue model. Then the evaluation model is further implemented in a model predictive-based control schematic, which targets to provide a continuous adaptive control to the user for regulating the muscle fatigue state. In experiments following characteristics of the developed device, evaluation approaches and control are demonstrated: --The developed device was successfully validated through a series of experiments demonstrating the capacity of the device to provide the target assistive functionality to the knee joint; --The GPR-based calibration method for the developed force sole shows a comparative reliable accuracy when compared with the model-based method; --The fault detection for the force sole can effectively detect the fault of the sensing elements, while the corresponding recovery method shows a significant improvement in the accuracy for extracting the COP and NGRF under fault conditions; --The proposed EMG-free muscle fatigue evaluation approach can provide continuous and reliable evaluation of the muscle fatigue state during the operation of the exoskeleton. --The proposed muscle fatigue-based model predictive control can provide adaptive control depending on the user's muscle fatigue state.

Le role des informations peripheriques dans la commande motrice a ete evalue au cours de mouvements simples de la mandibule. Ce systeme a comme particularite fondamentale une absence de fuseaux neuromusculaires dans les muscles abaisseurs, a l'inverse des muscles elevateurs. L'etude est basee sur l'enregistrement de l'activite electromyographique des muscles moteurs de l'articulation et des parametres cinematiques du mouvement. Dans un premier temps, il est montre que, tout comme pour les mouvements des membres, le snc dispose d'un repertoire de synergies motrices pour les mouvements mandibulaires elementaires. Le schema moteur au cours de ces mouvements est toutefois different de celui mis en evidence au cours des mouvements simples et rapides des membres. Cette difference d'organisation motrice est discutee en fonction des connexions entre les groupes musculaires antagonistes, et des particularites mecaniques du systeme. Outre leur role dans la mise en place des programmes moteurs, il est egalement montre que les informations sensorielles d'origine fusoriale sont preponderantes dans des taches de synchronisation motrice. Enfin, l'alteration des synergies motrices elementaires des muscles de la mandibule a ete evaluee dans l'ataxie de friedreich, maladie caracterisee entre autres symptomes par la dysarthrie. Les resultats suggerent qu'un deficit d'origine cerebelleuse pourrait dominer dans ces troubles des mouvements mandibulaires et etre a l'origine, tout au moins en partie, des desordres phonatoires observes dans l'ataxie de friedreich

Il campo della Creatività Computazionale prova a ottenere comportamenti creativi dagli elaboratori, per migliorare la comprensione di cosa regola la creatività e approfondire cosa sia possibile ottenere dai sistemi informatici. Una applicazione comune in questo campo è la composizione automatica di musica, in quanto questa attività notoriamente richiede creatività. In questa tesi descrivo i principali approcci alla generazione di musica, come anche alcune domande in questo campo che rimangono aperte. Nel lavoro qui descritto, mi focalizzo su due di queste, ovvero il ruolo delle emozioni e della socialità nella musica, e la struttura a lungo termine, cercando di capire il ruolo che questi due aspetti possono avere nella generazione di musica. Un gioco serio per l’interazione sociale è qui presentato. Il gioco permette a due giocatori di creare liberamente un ritmo usando due pad ritmici MIDI. Il programma riesce a seguire il tempo e il metro che i due suonano, aggiungendo musica alla loro interazione, aumentando così il valore estetico della loro esperienza di gioco e sociale. Questo gioco si dimostra un mezzo efficace per creare una esperienza coinvolgente per gli utenti. Viene presentato anche un sistema per l’analisi di strutture musicali. Questo algoritmo usa rappresentazioni ad hoc della musica, basate su rappresentazioni ad albero usate dai musicologi. Partendo da queste rappresentazioni di base che erano note in letteratura, l’algoritmo costruisce ulteriori astrazioni che sintetizzano gli aspetti strutturali di un intero brano, e poi di un intero corpus di brani. Alcune applicazioni di esempio vengono riportate, incluso un algoritmo per la generazione di musica che sfrutta queste reppresentazione e la Teoria dell’Informazione per creare nuovi brani che mostrino una struttura simile a quelle trovate nel corpus di esempio. Sebbene il metodo per la generazione di melodie usato in questi brani non è del tutto in grado di creare melodie realistiche, l’algoritmo riesce a generare strutture a lungo termine soddisfacenti, grazie alle rappresentazioni implementate.
The field of Computational Creativity tries to obtain creative behaviours from computers, to further the understanding of what regulates creativity and what is possible to obtain from computational systems. One common effort within this field is that to have computers write music, as this is an activity that is recognized to require creativity. In this thesis I review some of the main approaches to music generation as well as some questions that remain unanswered within the field. In the work described here, I focus on two of these, namely, the role of emotion and sociality in music, and long-term structure,trying to understand the role these two aspects can have in music generation. A serious game for social interaction is presented. This game allows two players to freely create a rhythm by interacting via MIDI drum pads. The software detects the tempo and meter they are playing, and adds a musical augmentation to their interaction, increasing the aesthetic value of their gaming and social experience. This is shown to be an effective way of creating a captivating experience for users. A system for the analysis of musical structure is also presented. This algorithm uses ad hoc representations of musical content, based on tree representations used by musicologists. Starting from the basic representations that were known in literature, this algorithm builds further abstracted representations that summarize the structural aspects of an entire piece and then of an entire corpus. Some example applications are shown, including an algorithm for music generation that leverages these representations and Information Theory concepts to create novel music that shows a structure similar to the ones found in the example corpus. While the method for generating the melodic material used for these novel pieces is not fully capable of generating realistic melodies, the algorithm manages to create satisfactory long-term structure, thanks to the implemented representations.

Objective: To explore patient factors that influence patients’ decisions about Neoadjuvant chemotherapy (NC) including patient unmet needs, challenges in decision making, and the role of family and caregiver involvement in treatment decision making. Our goal is to elucidate patient factors including their beliefs, values, expectations, worries about NC and MIBC, emotional challenges, and the importance of caregivers’ involvement in decision making and selection of treatment procedures.

Design: A qualitative study with semi-structured single interviews and focus groups. Atlas.ti software was used to analyze the qualitative data collected.

Settings: Icahn School of Medicine at Mount Sinai (ISMMS) and Northwell Health (NH), New York, USA.

Methods and Materials: 20 MIBC survivors from the ISMMS (n = 10) and NH (n = 10) were enrolled in this study from September 2015 and August 2016. 17 of them finished a semi-structured single interview and 3 of them joined a focus group interview to share their decision-making process about their treatment plans. Three ISMMS health care providers were interviewed to understand physicians’ standpoints on the challenges, barriers, and potential facilitators for MIBC patients to make a decision about their treatment plans.

Results: Patients decisions on whether or not to go through NC were influenced by three major parts: physician’s opinion or suggestion, family and spouses’ opinion, and patients’ own value and preferences. Sufficient information about treatment efficacy (i.e. treatment options, survival rates, experience sharing from former patients) would benefit patients’ decision-making process. Therefore ensuring information available for patients is important Results also showed that having a second opinion, experience shared by other patients, and caregivers’ help with information search helped address some of the patients’ reported informational needs. Physicians’ recommendations and advice were reported as influential factors in patients’ treatment decision making. Physicians’ challenges confirmed by interviewed patients suggested several areas await improvement including more help with daily activities from healthcare professionals, good bedside manners and professional figures, and opportunity to communicate with former patients.

Conclusions: Meeting patients’ informational and emotional needs is imperative to facilitate their decision-making process. Caregivers play significant roles in treatment decision making and recovery. Preparing caregivers for treatment decision making and patient care can maximize support received from the patient’s social network.

Diffusion Magnetic Resonance Imaging (dMRI) is a diagnostic technique able to provide in- vivo measures that are related to the microstructure of tissues. Thanks to the sensitivity to microstructural tissue changes, Diffusion Weighted Imaging (DWI) and derived metrics, as the Apparent Diffusion Coefficient (ADC) , became the gold standard for the detection of strokes and ischemia since the early 90‟s. In 1994 Basser and colleagues introduced Diffusion Tensor Imaging (DTI), the first quantification approach able to capture the anisotropy of the diffusion process in in-vivo biological tissues. Chapter II shows the results we obtained applying DTI to investigate white matter alterations of a population affected by Friedreich‟s Ataxia. After more than 20 years from its introduction, DTI is still widely applied. However, concerns about the limitations of the technique have been increasingly risen over-time, with particular reference to the lack of specificity of the model and the coexistence of tissues with multiple architectures. Additionally, the tensor model can be applied only to a range of “moderate” diffusion sensitizations, after which the presence of biological membranes becomes non-negligible and gives origin to phenomena of “non-Gaussian diffusion”, that violate the assumptions of the model. Chapter III and Chapter IV deal specifically with these limitations, addressing the problem with two different approaches and applications. Another popular technique to investigate the dMRI signal is Spherical Deconvolution (SD), that in Chapter IV is presented in a tissue specific formulation and applied to derive diffusivity metrics specific to white matter, gray matter and cerebrospinal fluid, both in healthy controls and in a patient affected by MS. Since the early days of dMRI, experiments have been performed not only in the brain but in several body districts, including the skeletal muscle. Back in 1986 Le Bihan et al. observed that the water flowing in the micro vascular network and in the vessels was contributing to the acquisition of data at very low diffusion sensitization, and proposed the “Intra- Voxel Incoherent Motion” (IVIM) model. IVIM can be seen either as a model to obtain measures of pseudo-diffusion, or as a technique to obtain perfusion free ADC measures, thus recognizing it as an artifact. Although dMRI and DTI were applied to the skeletal muscle since its early days, later evolutions as Diffusion Kurtosis Imaging have only recently been applied to the skeletal muscle to fit dMRI data acquired at strong diffusion sensitization. The concepts of IVIM and DKI are developed in Chapter V, where the effects of the first on DTI and DKI, as well as the relation between DTI and DKI metrics are investigated through simulations and MRI data of the calf. In line with the current dMRI literature, the first 5 chapters of this thesis depict the diffusion signal as a complex measure arising from multiple tissue components. Chapter VI investigates a multi-compartment pseudo-continuous deconvolution approach, a technique that does not require explicit modeling of the tissues. Finally, Chapter VII presents an overview of other research topics I have work on during the PhD.
L’Imaging mediante Risonanza Magnetica pesata in diffusione (diffusion-weighted Magnetic Resonance Imaging, dMRI) è una tecnica diagnostica in grado di fornire metriche correlate alla microstruttura dei tessuti. L’imaging pesato in diffusione (Diffusion Weigted Imaging, DWI) e tecniche derivate, quali il coefficiente di diffusione apparente (Apparent Diffusion Coefficient, ADC), in virtù della loro sensibilità ai cambiamenti microstrutturali nei tessuti, possono essere considerate lo stato dell’arte nella diagnosi di ictus ed ischemia cerebrale dai primi anni 90. Nel 1994 Basser e colleghi hanno introdotto l’imaging basato sul tensor di diffusione (Diffusion Tensor Imaging, DTI), la prima tecnica di quantificazione in grado di descrivere l’anisotropia del processo diffusivo nei tessuti biologici. Nel capitolo II sono illustrati i risultati di uno studio effettuato sulla sostanza bianca di pazienti affetti da Atassia di Friedreich con il modello DTI. Dopo oltre 20 anni dalla sua presentazione il modello DTI è ancora ampiamente utilizzato, tuttavia molteplici limitazioni sono state dimostrate spingendo verso un suo superamento, con particolare riferimento alla bassa specificità del modello e alla coesistenza di molteplici tessuti caratterizzati da architetture differenti. Inoltre, il modello tensoriale può essere applicato solo a dati acquisiti in uno specifico intervallo di pesature in diffusione, al di fuori del quale l’effetto delle membrane biologiche è non trascurabile originando fenomeni detti di diffusione non-gaussiana, che violano gli assunti del modello. Su questi argomenti vertono i capitoli III e IV, proponendo due differenti approcci per superare le limitazioni del DTI. Un’altra tecnica molto diffusa per analizzare il segnale dMRI è la deconvoluzione sferica (Spherical Deconvolution, SD), presentata nel capitolo IV in una formulazione tessuto-specifica ed applicata a soggetti di controllo e un paziente affetto da sclerosi multipla per derivare metriche specifiche a sostanza bianca, grigia e fluido cerebro-spinale. Storicamente, la dMRI è stata applicata non solo all’encefalo ma a diversi distretti periferici, incluso il muscolo scheletrico. Nel 1986 Le Bihan e colleghi osservarono come il segnale proveniente dall’acqua fluente nei vasi periferici e nella rete micro vascolare contribuissero al segnale dMRI acquisito con piccolo pesatura in diffusione, proponendo il modello “Intra- Voxel Incoherent Motion” (IVIM) per spiegarlo. IVIM può essere visto come un modello per ottenere misure di pseudo-diffusione o come una tecnica per ottenere misure di ADC ripulite dall’effetto della perfusione, categorizzandola come un artefatto. Nonostante la dMRI e il DTI siano stati applicati al muscolo scheletrico sin dalle origini, solo in tempi recenti evoluzioni quali il Diffusion Kurtosis Imaging sono state sperimentate con successo su dati muscolari acquisiti a forte pesatura in diffusione. I concetti di IVIM e DKI sono sviluppati nel capitolo V, dove gli effetti del primo su DTI e DKI, nonchè la relazione tra DTI e DKI sono analizzate mediante simulazioni e dati MRI della gamba inferiore. In linea con l’attuale letteratura dMRI, i primi 5 capitoli di questa tesi mostrano il segnale in diffusione come misura eterogenea. Il capitolo VI propone l’analisi di un metodo di deconvoluzione multi-compartimentale e pseudo-continuo, tecnica che non richiede di modellizzare esplicitamente il segnale dMRI. Infine, il capitolo VII presenta un sommario di altri argomenti di ricerca sviluppati nel corso degli studi di dottorato.

Le cœur joue un rôle vital dans le fonctionnement du corps humain. La fonction du cœur humain est de pomper le sang dans tout le corps, fournissant de l’oxygène et des nutriments aux tissus et en éliminant le dioxyde de carbone et d’autres déchets. Les maladies cardiovasculaires sont la première cause de décès dans le monde. Les maladies cardiaques sont principalement liées à un processus appelé athérosclérose. Ce processus cause un flux sanguin plus difficile à travers les artères et finalement il peut arrêter le flux sanguin. Il peut conduire à une crise cardiaque et un accident vasculaire cérébral. Un diagnostic précoce et précis des maladies cardiovasculaires joue un rôle important dans l’amélioration de la vie des personnes atteintes de maladies cardiaques. L’imagerie médicale est largement utilisée dans le diagnostic et le suivi des maladies cardiovasculaires. L’imagerie médicale est un processus de collecte d’informations sur un lieu d’intérêt dans le corps en utilisant une propriété caractéristique prédéfinie qui est affichée sous la forme d’une image. Les techniques d’imagerie permettent aux cliniciens et aux scientifiques de voir à l’intérieur du corps et de fournir une mine d’informations.Les progrès récents en imagerie médicale avec des contributions significatives de nombreux domaines de la science, tels que la physique médicale, la chimie, le génie électrique et informatique ont un impact important sur le diagnostic radiologique. Le développement de l’ingénierie et de l’informatique a donné la possibilité d’obtenir des images multidimensionnelles à haute résolution de la zone d'intérêt ciblée. Ce type d’images donne une information complexe pour analyser la structure et la fonction des organes pour un diagnostic assisté par ordinateur, plus précis ou destiné à développer ou diriger de nouvelles stratégies thérapeutiques. Le but de cette thèse est de développer un nouvel outil informatique qui aidera à obtenir des informations plus riches et pertinentes sur le myocarde à l'aide du traitement d'images.L’objectif principal de cette thèse est de développer une méthode complète combinant la tomographie par l'émission de positrons (PET) et l’enregistrement d’images par résonance magnétique (IRM). Pour faire face à la différence entre ces deux modalités, une méthode basée sur l’enregistrement des images présegmentées pour les deux modalités est proposée. La segmentation des images peut aider à extraire le myocarde de l’arrière-plan et se concentrer uniquement sur son enregistrement, sans l’impact de la structure environnante
The heart plays a vital role in the functioning of the human body. The function of the human heart is pumping blood throughout the body, supplying oxygen and nutrients to the tissues and removing carbon dioxide and other wastes. Cardiovascular diseases are the first cause of death worldwide. Heart diseases are mainly related to a process called atherosclerosis. This process caused harder blood flow through arteries and finally it can stop the blood flow. It can lead to heart attack and stroke. Early and accurate diagnosis of cardiovascular diseases plays an important role in improving the life of population afflicted heart diseases. Medical imaging is widely used in the diagnosis and monitoring of cardiovascular diseases. Medical imaging is a process of collecting information about a place of interest in the body using a predefined characteristic property that is displayed in the form of an image. Imaging techniques allow clinicians and scientist to see inside the body and provide a wealth of information.Recent advances in medical imaging with meaningful contributions from many fields of science, such us medical physic, chemistry, electrical and computer engineering, and computer science have a large impact on diagnostic radiology. The development of engineering and computer science has given the possibility to obtain high-resolution multidimensional images of the place of interest in the body. This kind of images gives a complex information to analyze the structure and function of the organs for computer-aided diagnosis, more accurate diagnosis or to develop or to direct new therapeutic strategies. The aim of this thesis is to develop a new method, which will allow to obtain more complex and accurate information about myocardial disease by using a computer science's and image processing methods.The main objective of this thesis is to develop a complete method allowing Positron Emission Tomography (PET) and Magnetic Resonance Images (MRI) registration of cardiac images. The main difficulties of the PET-MRI registration are the differences between these two modalities. To decrease these differences, the segmentation method were applied to PET and MRI images. Segmentation of the images can help to extract myocardium from the background and focus just on the registration of the myocardium without the impact of the structure around

Dans le secteur de loisirs il y a un nombre considérable d’enregistrements numériques musicaux produits, diffusés et échangés qui favorise la demande croisante de services intelligents de recherche de musique. La navigation par contenu devient cruciale pour permettre aux professionnels et également aux amateurs d’accéder facilement aux quantités de données musicales disponibles. Ce travail présente les nouveaux descripteurs de contenu musical et mesures de similarité qui permettent l’organisation automatique de données musicales (recherche par similarité, génération automatique des playlistes) ainsi que l’étiquetage (classification automatique en genres). Ce travail s’intéresse au problème de la construction des descripteurs du point de vue musical en complément des caractéristiques spectrales de bas-niveau. Plusieurs aspects d’analyse musicale, telles que l’analyse du signal où une nouvelle technique de transformation fréquentielle à résolution variable est proposée et décrite. Le traitement de niveau plus haut touche aux aspects de l’extraction des connaissances musicales. Cette thèse présente les algorithmes de détection de coups (beats) et d’extraction de fréquences fondamentales multiples. Les deux algorithmes sont basés sur la transformation à résolution variable proposée. Les informations issues de ces algorithmes sont utilisées dans la construction des descripteurs musicaux, représentés sous forme d’histogrammes (nouvel histogramme rythmique 2D qui permet d’estimer directement le tempo, et les histogrammes de succession et profil de notes). Deux applications majeures qui utilisent les caractéristiques mentionnées sont décrits et évaluées dans cette thèse
As a major product for entertainment, there is a huge amount of digital musical content produced, broadcasted, distributed and exchanged. There is a rising demand for content-based music search services. Similarity-based music navigation is becoming crucial for enabling easy access to the evergrowing amount of digital music available to professionals and amateurs alike. This work presents new musical content descriptors and similarity measures which allow automatic musical content organizing (search by similarity, automatic playlist generating) and labeling (automatic genre classification). The work considers the problem of content descriptor building from the musical point of view in complement of low-level spectral similarity measures. Several aspects of music analysis are considered such as music signal analysis where a novel variable resolution transform is presented and described. Higher level processing touches upon the musical knowledge extraction. The thesis presents algorithms of beat detection and multiple fundamental frequency estimation which are based on the variable resolution transform. The information issued from these algorithms is then used for building musical descriptors, represented in form of histograms (novel 2D beat histogram which enables a direct tempo estimation, note succession and note profile histograms etc. ). Two major music information retrieval applications, namely music genre classification and music retrieval by similarity, which use aforementioned musical features are described and evaluated in this thesis

La musique est présente dans tous les métrages produits par les industries audiovisuelles et cinématographiques. Elle peut préexister au film, avoir pour destination l’industrie traditionnelle de la musique ou être spécialement prévue pour être synchronisée à des images. Dans ce cas, la musique recherchée est généralement perdue dans des bases de media musicaux disposant de leurs propres Systèmes de Recherche d’Information (SRI). Alors que dans le monde anglo-saxon la profession de Music Supervisor (MS) permet d’allouer à cette tâche un vrai spécialiste de l’utilisation de ces systèmes, il n’en est pas de même dans les productions d’Europe continentale. Des MS per accidens s’y trouvent devoir traduire l’humeur musicale désirée en requêtes pour ces SRI qu’ils interrogent parfois pour la première fois. Dans l’optique de rendre cette tâche moins ardue pour des non-spécialistes de la question musicale, nous avons exploré deux voies. L’une envisage la proxémie sémantique pour le nettoyage des nuages de tags mêlant taxinomies et folksonomies, afin de tracer de nouvelles reliances entre media ou écarter des qualifications qualitativement aberrantes. La seconde voie explorée consiste à trouver une manière d’éviter la verbalisation. Après un état de l’art des solutions existantes pour la qualification de l’humeur ou, plus largement, des états émotionnels, solutions développées pour les besoins de nombreuses autres disciplines (psychologie expérimentale, sciences de l’éducation, marketing, etc.), l’analyse conduit au prototypage du dispositif SYM (Spot Your Mood). Par pointage sur un espace de Valence-Arousal, SYM permet de substituer des coordonnées spatiales abstraites relativement intuitives aux qualifiants verbaux. Au-delà du champ musical, SYM a aussi fait l’objet d’autres validations dans la qualification de l’expérience d’utilisateurs ou de spectateurs, ainsi que pour l’aide à l’explicitation de l’humeur chez des publics souffrant de divers handicaps empêchant la verbalisation ou la communication
The music is part of all films produced within the audiovisual and cinematographic industries. It can preexist the film, be destined to be part of the traditional music industry or to be especially synchronized with pictures. In this case, the music excerpt sought after is usually lost in endless media bases providing their own Information Retrieval Systems (IRS). Whereas in the Anglo-Saxon world Music Supervisors (MS) are genuine experts in the use of these systems, it is not the same across Continental Europe. Per accidens MS are asked to translate the desired musical mood into queries for these IRSs, systems that they are sometimes using for the first time. In order to make this task easier for these non-specialists, we explored two approaches. The first led us to explore the possibilities offered by semantic proxemics in order to clean tag clouds mixing taxonomies and folksonomies to trace new relationships between media or to reject qualitatively aberrant qualifications. The second consisted in finding a way to do so without verbalisation. After a review of extant solutions developed for the needs of many other disciplines (experimental psychology, sciences of education, marketing, etc.), we managed to prototype SYM (Spot Your Mood). Instead of relying on verbalisations, SYM makes it possible to provide relatively intuitive spatial coordinates out of a valence-arousal space. Also, SYM has been put to use for many other adaptations in order to qualify user or spectator experience as well as to help elicit the mood with audiences suffering from various verbal or communicational impairments

The force produced by a specific muscle cannot be measured and what is measured is the total force provided by all the active muscles acting on a joint, including agonists and antagonists. The first part of this work (chapter 3) addresses the issue of load sharing by proposing two possible approaches and testing them. The second part (chapter 4 and 5) addresses two applications of surface EMG focusing on the study of a) muscle relaxation associated to Yoga sessions and b) the activation of muscle of the back and shoulder of musicians playing string instruments (violin, viola and cello). In both parts the element of innovation is the use of two dimensional electrode arrays and of techniques based on EMG Imaging. The objectives of this work are presented and explained in chapter 1 while the basic concepts of surface EMG are summarized in chapter 2. Different EMG-based muscle force models found in the literature are explained and discussed. Two renowned amplitude indicators in surface EMG (sEMG) studies are the average rectified value (ARV) and the root mean square (RMS). These two amplitude indicators are computed over a defined time window of the recorded signals to represent the muscle activity. The advantages and disadvantages of RMS and ARV are compared and discussed for a simple sinusoid as well as for more complex signals (simulated motor unit action potential detected by high density electrode grid). The results show that RMS is more robust to the sampling frequency than ARV. In this thesis, starting from the simulation of a single fiber and of a group of fibers (motor unit), it is shown that inter electrode distance (IED) greater than10 mm causes aliasing. Aliasing is a source of error in sEMG map interpretation or decisions that are made by automatic algorithms such as those providing image segmentation for the identifications of regions of interest. Chapter 2 discusses three segmentation algorithms (K-means, h-dome, watershed) and compares them in order to find the most suitable method. Results reveal that among the three mentioned algorithms, watershed provides most accurate segmentation for the simulated ARV maps. Chapter 3 presents a mathematical model that is associated to the monotonic Force-EMG relation. A possible non-linear relationship between the EMG and force or torque is presented. A system of "M" equations is obtained by performing "M" measurements at "M" different force levels in isometric conditions. The solutions of such system of equations are the values for each muscles. Two different approaches were investigated for finding the solutions of the system, which are: a) Analytical-Graphical Approach (AGA) and b) Numerical Approach (NA) consisting of error minimization (between the total estimated and measured force) applying optimization algorithms. The AGA was used to find the model parameters of each muscle contributing to the force production on a joint by finding the intersection of those surfaces that can be obtained from sequential substitutions of the model parameters in the equations corresponding to each contraction level. In simulation studies, the AGA graphically shows that there is more than one solution to the load sharing problem even for the simplest theoretical case (i.e. a joint spanned by only two muscles). The second approach, based on minimization of the mean square error between the measured and the total estimated force or torque (with "N" muscles involved) provides an estimate of the model parameters that in turn provides the force contributions of the individual muscles. The optimization algorithms can find the solutions of our system made of non-linear equations (see chapter 3). Starting from different point (initial conditions), different solutions can be found, as predicted by the AGA approach for the two-muscle case. The main conclusion of this study is that the load sharing strategy is not unique. Chapter 4 discusses the application of surface electromyography to a single case study of Yoga relaxation to show the feasibility of measurements. The effect of yoga relaxation on muscle activity (sEMG amplitude), as well as on mean and median frequencies and muscle fiber's conduction Velocity, is discussed in this chapter. No changes in the sEMG activity pattern distribution were found for the same task performed before and after applying yoga relaxation technique. However, myoelectric manifestations of fatigue were smaller after relaxation and returned to the normal pattern after the recovery phase from relaxation. Further studies are justified. Chapter 5 describes results and discusses the spatial distribution of muscle activity over the Trapezius and Erector Spinae muscles of musicians playing string instruments. In chapter 5, the effect of backrest support in sitting position during playing cello, viola, and violin on the muscle activity index of upper and lower Trapezius muscle of the bowing arm, upper Trapezius muscle of non-bowing arm, left and right Erector Spinae muscles is investigated. Two professional players (one cello and one viola) and five student players (one cello, three violin and one viola) participated in this study. The muscle activity index (MAI) was defined as the spatial average of RMS values of the muscle active region detected by watershed segmentation for Trapezius muscles (left and right), and thresholding technique (70% of the maximum value) for left and right Erector Spinae muscles. It was found that the MAI is string (note) dependent. Statistical difference (p < 0:05) between the MAIs of left Erector Spinae muscle during playing with and without backrest support was observed in four (out of five) student players. No statistical differences were observed on the muscle activity of Trapezius (bowing and no-bowing arms) during playing with and without backrest support in different types of bowing for all musicians. In conclusion, this work addresses a) the issue of spatial sampling and segmentation of sEMG using 2D electrode arrays, b) two possible approaches to the load-sharing issue, c) a single-case study of Yoga relaxation and d) the distribution of muscle activity above the Trapezius and Erector Spinae muscles of musicians playing string instruments. Previously unavailable knowledge has been achieved in all these four studies.

By recording surface EMG using standard bipolar EMG, previous studies have demonstrated that elderlies tend to activate their postural muscles during standing with a higher degree of activity and for a prolonged duration compared with young adults. The EMG biofeedback technique has been widely used to reduce the excessive level of muscle activity in different fields, e.g. the prevention and reduction of low back pain. In this view, EMG biofeedback could be a potential tool to assist aged subjects in reducing the excessive muscle activity during standing balance. However, whether the greater, prolonged activation observed locally in the muscles of aged subjects reflects the activation of the entire muscle is still an open question. It is possible that differences in the activation of postural muscles with aging are more or less expressive than previously appreciated. This thesis aimed at obtaining new insights into the rationale and the effects of the use of EMG biofeedback for the improvement of muscle efficiency during standing in the elderly. It was evaluated whether muscle activation during standing differ between young and aged subjects through a sophisticated detection system for the acquisition of surface EMGs from multiple regions of a single muscle (i.e., high-density surface EMG). Before to test this hypothesis, a methodological issue was addressed to verify whether high-density surface EMG is selective enough to detect during standing: (a) different activation between ankle muscles, as observed with other techniques (intramuscular electromyography); (b) variations in the activity within ankle muscles (i.e., soleus muscles). The results of this methodological study revealed that the medial portion of soleus muscle was activated continuously compared to the lateral portion of soleus and medial gastrocnemius, which were activated intermittently. These results suggest high-density surface EMGs can be used to discriminate the activity between ankle muscles (i.e., medial gastrocnemius and soleus) and muscle activity sampled from different regions of a single muscle (i.e., soleus) can provide estimates more representative of muscle activity during standing. High-density surface EMG was therefore used to assess muscle activity between young and aged subjects during standing. Key results indicate that during standing: (a) tibialis anterior and medial gastrocnemius muscles were active for a longer duration in aged than young subjects; (b) a greater proportion of medial gastrocnemius volume was active in aged individuals. Collectively, these results corroborate previous evidence that elderlies tend to stand with a greater muscle effort than young subjects. Thus, the well-documented attenuation effect of EMG biofeedback on muscle activity may extend to the control of human standing posture with aging. This thesis addressed additional issues which could be relevant to provide more representative EMGs of muscle activity to the subject through EMG biofeedback and to prove the attenuation effect of EMG biofeedback on the activity of lower limb muscles during standing. The following two questions were addressed: i) should EMGs be sampled from both lower limbs to provide more representative information about calf muscles activity? It was observed differences in muscle activity between left and right ankle muscle while young subjects stood at ease. These results indicate muscle activity should be sampled from the ankle muscles of both legs to avoid a biased recording and feedback of muscle activity during standing. ii) Is the attenuation effect of EMG biofeedback on the ankle muscles activity generalized to – or compensated by – other muscles during standing? These findings revealed the attenuation effect of EMG-audio feedback on ankle muscles is not compensated by other lower limbs muscles not included for the feedback. Therefore, the EMG biofeedback may be a promising technique to assist individuals in more efficiently controlling lower limbs muscles during standing. If the short-term, attenuation, effect of EMG-audio feedback on ankle muscles’ activity in young individuals observed here is generalized to other populations (e.g., the elderly) and retained after training, then, improvement in postural muscle efficiency may contribute significantly to an ability to maintain standing balance, to respond to unexpected perturbation, standing on narrow stances and walking; with potential implication for the prevention of falls.

Nos travaux portent sur les mécanismes nerveux qui sous-tendent la perception de notre corps en mouvement. Chacune de nos actions est caractérisée par l'intention motrice qui en est à l'origine et par le feedback multi-modalitaire que son exécution évoque. Dans ce cadre, nous avons formalisé le codage proprioceptif du mouvement et nous avons exploré deux types d'images motrices et leur interaction perceptive à travers l'étude des mouvements imaginés et des illusions kinesthésiques.Sur la base de données neurosensorielles, nous avons contribué à développer une méthode « propriomimétique » inspirée du modèle vectoriel de population permettant de prédire les patterns proprioceptifs afférents évoqués lors de mouvements complexes. Ces patterns modélisés ont ensuite été utilisés pour activer les voies de la proprioception musculaire chez des sujets immobiles grâce à la méthode des vibrations tendineuses
This thesis deals with the neural mechanisms involved in body movement perception. We contributed to developing a "propriomimetic" method based on neurosensory data and drawn from the population vector model for predicting the afferent proprioceptive patterns evoked during complex movements.The main contribution of this work is that it extends the scope of the method to include the whole set of possible human actions by showing that it can be used to accurately predict the proprioceptive patterns and to generate the kinaesthetic experiences associated with movements involving one or more joints, performed in 2- or 3-dimensional space, regardless of which muscles are involved.Other motor images were intentionally generated by the subjects in subsequent experiments and combined with those imposed by the vibratory stimulation. The results obtained here show that the images of both kinds were completely integrated when evoked simultaneously and that they gave rise to a unique and original perceptual experience, in which their spatio-temporal characteristics were combined. From the theoretical point of view, these findings confirm that proprioceptive afferents play an important role, along with the motor intention itself, in the elaboration of kinaesthetic perception. From the clinical point of view, the possibility of evoking motor images quite "naturally" using the method presented here to activate the peripheral receptors and/or the command centers, and thus the whole sensorimotor loop in the absence of any real movements, opens some promising perspectives for rehabilitation purposes

E' stato affrontato lo studio di un'algoritmo per l'addestramento di reti neurali sul riconoscimento di genere sul supercalcolatore ENEA-CRESCO, al fine di poter essere sfruttato per il progetto Share Art, un programma ENEA. Share Art consiste in un dispositivo che permette, attraverso una telecamera ad infrarossi, di rilevare automaticamente i volti che guardano in direzione della telecamera, acquisendo contestualmente una serie di informazioni relative all’osservazione delle opere d’arte, come ad esempio il tempo di permanenza, la distanza, il percorso compiuto. L'addestramento di una rete neurale su un'architettura diversa dalla nostra, quale è CRESCO, ha previsto una mirata predisposizione dell'ambiente di sviluppo, così da isolare le componenti necessarie utilizzate dall'algoritmo. Procedura fondamentale è stata quindi quella di pacchettizzare in un container il tutto, renderlo accessibile all'applicazione, e poter utilizzare ciò che l'applicazione necessitava per essere eseguita senza i permessi di amministratore di sistema. Verranno mostrati i passi che sono stati seguiti per la creazione dell'ambiente di sviluppo, del container Docker e della predisposizione delle dipendenze necessarie, come ad esempio le librerie Tensorflow e Keras. Alla base del sistema Share Art le telecamere generano un flusso di dati che è l'input dell'analisi che viene eseguita direttamente sul Raspberry per verificare ad esempio il corretto utilizzo della mascherina e il rispetto del distanziamento sociale. I dispositivi Share Art installati presso il museo, generano quindi dati importanti della ricerca in ambito museale. Sarà quindi affrontato il tema della gestione dei dati e del loro trasferimento riformulando un programma che apporti scalabilità ad un sistema già esistente. Si è scelto di utilizzare un protocollo standard dell'IoT, MQTT, il tutto supportato da un linguaggio sviluppato presso l'Università di Bologna, Jolie.

Les psychologues (comme A. Damasio, K. R. Scherer, P. Ekman) ont montré que l’émotion est un élément essentiel dans la prise de décision, dans l’évolution des capacités d’apprentissage et de création, dans l’interaction sociale. Il est donc naturel de s'intéresser à l'expression d'émotions dans le cadre de l'interaction homme-machine. Nous avons proposé dans un premier temps le modèle GRACE, modèle générique des émotions pour les applications computationnelles. Nous nous sommes basés en particulier sur la théorie psychologique de K. R. Scherer, qui cherche à produire une théorie des processus émotionnels qui soit modélisable et calculable. La pertinence de notre modèle a été vérifiée et validée via une comparaison avec différents modèles computationnels existants. Si le modèle GRACE est générique, nous nous sommes attachés à montrer qu’il pouvait s’instancier dans un contexte particulier, en l’occurrence l’interaction homme-robot utilisant la modalité musicale. Nous nous sommes intéressés pour cela d’une part à la conception d’un module d’analyse du contenu émotionnel d’une séquence musicale, d’autre part à la conception de mouvements émotionnellement expressifs pour un robot mobile. Du point de vue de l’analyse musicale, la contribution principale de la thèse porte sur la proposition d’un ensemble réduit d’indicateurs musicaux et sur la validation du module d’analyse sur une base de données de grande taille conçue par un expert en musicologie. Du point de vue de la robotique, nous avons pu montrer expérimentalement qu’un robot avec des capacités expressives très limitées (déplacements, mouvements de caméra) pouvait néanmoins exprimer de manière satisfaisante un ensemble réduit d’émotions simples (joie, colère, tristesse, sérénité)
Emotion, as psychologists argue (like A. Damasio, K. R. Scherer, P. Ekman), is an essential factor for human beings in making decision, learning, inventing things, and interacting with others. Based on this statement, researchers in Human-Machine Interaction have been interested in adding emotional abilities to their applications. With the same goal of studying emotional abilities, we propose, in our work, a model of emotions named GRACE, which helps in modelling emotions in computational applications. We based our model on the work of psychologist Klaus R. Scherer, who intensively searches to form a generic model of emotion applicable to computational domain (like informatics, robotics, etc.). We demonstrate the pertinence of our model by comparing it to other existing models of emotions in the field of informatics and robotics. In this thesis, we also worked on the instantiation of GRACE, in particular the components Cognitive Interpretation and Expression. These two components have been developed to be applied in the context of interacting with users using music. To develop Cognitive Interpretation, we worked on the extraction of emotional content in musical excerpts. Our contribution consists in proposing a reduced number of musical features to efficiently extract the emotional content in music, and in validating them via a learning system with a large database designed by a musicologist. For Expression, we have worked on the design of emotional moves of a mobile robot. Through very limited moves (moves in space, camera moves), we have shown that with dance-inspired motions, the robot could efficiently convey basic emotions (i.e. happiness, sadness, anger, serenity) to people

La realtà aumentata, unitamente a quella mista, stanno rapidamente prendendo pieno all'interno di molti aspetti della vita umana. Scopo di questo lavoro è di analizzare tecnologie e tecniche esistenti al fine di applicarle ad un caso reale, la rilevazione e la sovrapposizione di un oggetto digitale tridimensionale ad uno presente in un museo.

This thesis focuses on the functional study of the Central Nervous System (CNS) with non-invasive techniques. Two different aspects are investigated: nonlinear aspects of the cerebrovascular system, and the muscle synergies model for motor control strategies. The main objective is to propose novel protocols, post-processing procedures or indices to enhance the analysis of cerebrovascular system and human motion analysis with noninvasive devices or wearable sensors in clinics and rehabilitation. We investigated cerebrovascular system with Near-infrared Spectroscopy (NIRS), a technique measuring blood oxygenation at the level of microcirculation, whose modification reflects cerebrovascular response to neuronal activation. NIRS signal was analyzed with nonlinear methods, because some physiological systems, such as neurovascular coupling, are characterized by nonlinearity. We adopted Empirical Mode Decomposition (EMD) to decompose signal into a finite number of simple functions, called Intrinsic Mode Functions (IMF). For each IMF, we computed entropy-based features to characterize signal complexity and variability. Nonlinear features of the cerebrovascular response were employed to characterize two treatments. Firstly, we administered a psychotherapy called eye movement desensitization and reprocessing (EMDR) to two groups of patients. The first group performed therapy with eye movements, the second without. NIRS analysis with EMD and entropy-based features revealed a different cerebrovascular pattern between the two groups, that may indicate the efficacy of the psychotherapy when administered with eye movements. Secondly, we administered ozone autohemotherapy to two groups of subjects: a control group of healthy subjects and a group of patients suffering by multiple sclerosis (MS). We monitored the microcirculation with NIRS from oxygen-ozone injection up 1.5 hours after therapy, and 24 hours after therapy. We observed that, after 1.5 hours after the ozonetherapy, oxygenation levels improved in both groups, that may indicate that ozonetherapy reduced oxidative stress level in MS patients. Furthermore, we observed that, after ozonetherapy, autoregulation improved in both groups, and that the beneficial effects of ozonetherapy persisted up to 24 hours after the treatment in MS patients. Due to the complexity of musculoskeletal system, CNS adopts strategies to efficiently control the execution of motor tasks. A model of motor control are muscle synergies, defined as functional groups of muscles recruited by a unique central command. Human locomotion was the object of investigation, due to its importance for daily life and the cyclicity of the movement. Firstly, by exploiting features provided from statistical gait analysis, we investigated consistency of muscle synergies. We demonstrated that synergies are highly repeatable within-subjects, reinforcing the hypothesis of modular control in motor performance. Secondly, in locomotion, we distinguish principal from secondary activations of electromyography. Principal activations are necessary for the generation of the movement. Secondary activations generate supplement movements, for instance slight balance correction. We investigated the difference in the motor control strategies underlying muscle synergies of principal (PS) and secondary (SS) activations. We found that PS are constituted by a few modules with many muscles each, whereas SS are described by more modules than PS with one or two muscles each. Furthermore, amplitude of activation signals of PS is higher than SS. Finally, muscle synergies were adopted to investigate the efficacy of rehabilitation of stiffed-leg walking in lower back pain (LBP). We recruited a group of patients suffering from non-specific LBP stiffening the leg at initial contact. Muscle synergies during gait were extracted before and after rehabilitation. Our results showed that muscles recruitment and consistency of synergies improved after the treatment, showing that the rehabilitation may affect motor control strategies.

The freshwater pearl mussel (Margaritifera margaritifera) has declined dramatically throughout its range and is faced with recruitment problems in most of the streams where populations still persist. Human activities, such as forestry and agriculture, are thought to be the main reason for these problems. In this study, spatial information on landscape features along 38 streams with known recruitment status were analyzed in an effort to determine if GIS-tools could be used to identify streams with successful recruitment. Differences in the distribution of various landscape features, measured in stream corridors of 50 and 150 m, and differences in several water chemical factors between streams with and without recruitment were investigated. The distribution of landscape features was also compared with host fish (brown trout) density and any statistically significant water chemical factor.  Both mussel recruitment and trout density were found to be negatively related to clear-cuts, and mussel recruitment was also negatively related to high water color, which has been shown to be correlated with high nutrient content, one of several adverse effects of clear-cutting close to streams. Recruitment was expected to be negatively affected by roads, but no such relationship could be found. Instead, mussel recruitment was found to be positively related to the number of road crossings per kilometer, but the strength of this relationship was questionable. In addition, even though it was somewhat unclear, high proportions of lakes and ponds were found to be positive for both recruitment and high trout density. The results of the study indicate that GIS-tools can be used to find landscape features that affect recruitment of freshwater pearl mussel and they support the belief that forestry activities are an important cause for the decline of the species in Sweden. In addition, the results indicate that leaving protective zones of forest between streams and clear-cuts can be a possible conservation method for the freshwater pearl mussel.

Flodpärlmusslan (Margaritifera margaritifera) har minskat kraftigt i hela sitt utbredningsområde och i de flesta vattendrag där populationer finns kvar sker ingen rekrytering. Mänskliga aktiviteter, exempelvis skogsbruk och jordbruk, anses vara huvudorsaken för dessa problem. Syftet med den här studien var att undersöka om GIS-verktyg kan användas för att identifiera vattendrag med fungerande rekrytering. För att uppnå det analyserades geografisk information om landskapet längs 38 vattendrag med känd rekryteringsstatus. Skillnader i fördelningen av olika landskapsföreteelser, uppmätt i 50 och 150 meters buffertzoner, och skillnader i flera vattenkemiska faktorer mellan vattendrag med eller utan rekrytering undersöktes. Dessutom undersöktes även om det fanns något förhållande mellan landskapsföreteelser och tätheten av värdfisk (öring) och mellan landskapsföreteelser och signifikanta vattenkemiska faktorer. Både rekrytering av flodpärlmusslor och öringstäthet var negativt relaterad till kalhyggen. Rekrytering var också negativt relaterad till hög vattenfärg, som har visats vara korrelerat med högt näringsinnehåll, en av flera effekter kalhyggen nära vattendrag har. Rekrytering förväntades påverkas negativt av vägar, men inget negativt förhållande hittades. Istället hittades ett positivt förhållande mellan rekrytering och antalet vägkorsningar per kilometer, men styrkan av det förhållandet var ifrågasättbar. Dessutom var, om än något otydligt, andelen sjö och damm positivt för både rekrytering och hög öringtäthet. Studien indikerar att GIS-verktyg kan användas för att hitta landskapsföreteelser som påverkar rekryteringen av flodpärlmusslor och den stödjer bedömningen att skogsbruk är en av de viktigaste anledningarna till artens nedgång i Sverige. Dessutom indikerar resultaten att skyddszoner mellan kalhyggen och vattendrag kan vara en möjlig skyddsåtgärd för flodpärlmusslan.

La tesi descrive lo sviluppo di un'applicazione web per comporre musica tramite la tecnica del "live looping" che fornisce anche la possibilità di effettuare lo streaming di ciò che si crea in tempo reale e in maniera peer-to-peer. L'applicazione in oggetto (chiamata WebLooper) fa uso di due tecnologie web emergenti in ambito multimediale: Web Audio e WebRTC, attualmente in attesa di diventare standard W3C.

Le commentaire d’enregistrement à l’écrit mobilise de nombreuses compétences liées à la perception et à la rédaction. Ce travail, conçu avec la volonté d’un rapprochement interdisciplinaire, tente de montrer l’influence de la révision autonome de l’écrit ainsi que celle des œuvres écoutées sur des textes d’élèves de troisième. Les bases théoriques sont celles de la psychologie de l’audition et de la linguistique. L’effet dominant de la maîtrise de la langue s’estompe à la faveur d’une description systématique et globale du vocabulaire recueilli. Cette thèse rassemble des outils d’analyse de contenu originaux élaborés en fonction du corpus. Les apports sont donc essentiellement d’ordre méthodologique. Les indicateurs retenus permettent d’évaluer les écarts de densité informative selon l’œuvre écoutée et selon le rang d’écriture du commentaire. La synthèse des résultats montre les possibilités de détermination d’une typologie des écrits et révèle certains aspects stylistiques et formels des œuvres supports par l’observation raisonnée des zones d’intensification lexicale
Writing a commentary after a listening exercise mobilizes several skills linked to perception and drafting. This thesis, aiming at an interdisciplinary approach, attempts to describe the influence of autonomous revision and of the piece of music heard on commentaries written by pupils in the fourth form. Its theoretical foundations are found in audition psychology and linguistics. The primary role attributed to language skills gives way to a systematic and holistic description of the compiled vocabulary. For this thesis, original content analysis tools have been designed, based on the corpus. Its contribution is thus essentially methodological. The selected indicators enable us to assess the disparities in informative density according to the piece heard and the drafting register of the commentary. Based on a synthesis of the results obtained, it is thus possible to establish a typology of written commentaries and bring to light some stylistic and formal aspects of the selected pieces of music through a reasoned analysis of lexical intensification areas

Questa trattazione verterà sull'analisi di un sistema di gioco tra due individui, tra loro interagenti mediante handeld device, in un contesto immersivo posto in un luogo chiuso di natura museale. L'intera struttura, sviluppata nell'ambito del Corso di Sistemi ed Applicazioni Multimediali nell'anno 2010-2011, vuole porre un'innovazione rispetto agli scenari esistenti: laddove un'interazione museale richieda spesso la staticità dell'utilizzatore del servizio, il mezzo del gioco richiede invece la manipolazione e il movimento all'interno del luogo, garantendo un coinvolgimento maggiormente accentuato rispetto alla normale fruizione dei contenuti. Inoltre, la scelta alla base della realizzazione realizzazione si affida allo strumento multimediale e sempre più diffuso costituito dalle tecniche di Realtà Aumentata, che appongono aggiunte digitali all'ambiente circostante e permettono di modellare strutture d'informazione e manipolazione percepibili solo tramite apposite interfacce, ma pienamente inserite in un tessuto reale. La componente innovativa in questo schema è però portata da un'ulteriore aspetto, quello della cooper- azione di due dispositivi in contemporanea attori dell'interazione che, attraverso l'azione sulle informazioni esistenti, consentono di produrre modifiche rilevabili da ciascuno dei coinvolti.

Ce travail a pour but principal de répondre à certaines difficultés de notation rencontrées par les compositeurs. Dans un premier temps, nous nous se proposons de montrer les limites du système d’écriture musicale occidentale (SEMO) dans la notation de certains éléments de la musique et, dans un deuxième temps, de montrer les avantages qu’offre l’utilisation de l’écriture multimédia (EM). Les résultats de ce travail, obtenus à partir de six « expériences-projets de composition », montrent l’efficacité de l’EM : un ensemble de procédés qui permettent l'utilisation simultanée de plusieurs modes de représentation de l'information (tels que textes, sons, images fixes ou animées, entre autres), servant à noter un message afin de pouvoir le conserver et le transmettre du compositeur à l’interprète. Dans le cadre de la composition musicale, l’EM, telle que nous l’avons employée ici, s’est montrée efficace pour noter les éléments suivants : le timbre, des nouveaux modes de jeu, des nouvelles techniques vocales, des instructions pour l’emploi des logiciels ; et aussi l’EM s’est montrée efficace pour l’intégration des éléments suivants dans les œuvres de musique mixte : le jeu d’un musicien de tradition orale, des modes de jeu tirés des musiques de tradition orale et exécutés par un musicien de tradition écrite, une langue tonale (et l’expressivité liée aux genres de musique de tradition orale dans les œuvres écrites. À l’ensemble de notations multimédia déjà existantes et disponibles pour la composition musicale, nous ajoutons deux types de notation qui se sont révélés efficaces dans cette recherche : la notation auditive et la notation d’un savoir-faire au moyen de la vidéo
The main goal of this work is to solve some of the difficulties that composers encounter when notating music. Firstly, we describe how the Western musical notation (WMN) is limited when attempting to write specific musical elements. Secondly, we show the possible advantages that multimedia writing (MW) could offer on the notation of such elements. To address these issues, we used MW in six “experiments/composition projects” that were conceived to answer specific notation questions. The results obtained thereof allowed us to demonstrate the efficacy of MW for overcoming current limitations in music notation. More specifically, MW constitutes a group of procedures that allows to simultaneously represent information in different ways. This information could be either text, sounds, still or moving images, among others. Such procedures can be used to note down a message to further save it and transfer it from the composer to the performer. In the context of our experimental paradigms, MW has proven to be efficient for: the notation of timber, the integration of musicians from oral tradition in mixed music works, the incorporation of instrumental techniques drawn from oral tradition music into written music, the integration of a tonal language in a music score, the notation of new instrumental and vocal techniques, the guidance at using new software, and the incorporation of expressiveness associated to music styles of oral tradition into written works. We consider that two types of notation that proved to be efficient in this research could be added to the body of already existing MW, namely auditive notation and the notation of a know-how by the means of video

Les partitions musicales sont des artefacts d’une grande complexité sémiotique visant à transmettre toutes les informations relatives à la structure et au contenu d’une œuvre musicale de manière compacte, lisible et cohérente. De nos jours, la musique est la plupart du temps diffusée sous forme enregistrée. Un système capable de transformer automatiquement une interprétation musicale en partition serait donc extrêmement bénéfique pour les interprètes et les musicologues. Cette tâche, appelée “automatic music transcription” (AMT), comprend plusieurs sous-tâches qui impliquent notamment la transformation de représentations intermédiaires comme le MIDI, quantifié ou non. Nous d´défendons dans cette thèse l’idée qu’un modèle robuste et bien structure des informations contenues dans une partition musicale aiderait au développement et à l’évaluation de systèmes AMT. En particulier, nous préconisons une s´séparation claire entre le contenu musical encodé dans la partition et l’ensemble des signes et des règles de notation utilisés pour représenter graphiquement ce contenu
A musical score is a complex semiotic object that excels at conveying musical information in a human readable format. Nowadays, a lot of music is available exclusively as recorded performances, so systems which can automatically transform those performances into musical scores would be extremely beneficial for performers and musicologists. This task, called automatic music transcription (AMT), comprises a large number of subtasks which generically transform the performance input into higher level representations, such as unquantized and quantized MIDI files. We believe that a clear model of the information contained in a musical score would help the development and the evaluation of AMT systems. In particular we advocate for a clear separation between the music which the score encodes and the set of notation symbols which is employed to represent it

Ce travail rapporte des experiences realisees sur des chats anesthesies pour examiner, par enregistrement intracellulaire, les effets des afferences proprioceptives activees par la contraction d'un muscle extenseur de la cheville sur les motoneurones spinaux lombaires et sur les cellules du faisceau spino-cerebelleux dorsal (dsct). Dans les motoneurones homonymes et synergistes, la contraction prolongee du gastrocnemien median (gm) provoque une inhibition, due a l'action des afferences ib, dont l'efficacite diminue rapidement, car les informations transmises par ces afferences sont filtrees par un mecanisme d'inhibition presynaptique. Dans les motoneurones innervant des muscles antagonistes et des muscles agissant sur le genou, la contraction prolongee du gm evoque aussi une inhibition a decroissance rapide due a l'action d'afferences du groupe ii innervant des terminaisons fusoriales secondaires et/ou des mecanorecepteurs non specifiques actives par la contraction. Le retro-controle inhibiteur des motoneurones par la proprioception musculaire privilegie l'information sur le developpement de la force, parametre dynamique de la contraction. Pendant la contraction prolongee du gm, certains neurones du dsct presentent, comme les motoneurones, des inhibitions a decroissance rapide. Mais, d'autres sont excites pendant toute la duree de la contraction et transmettent au cervelet et au cerveau des informations sur la duree de la force, parametre statique de la contraction

Il est raisonnable de considérer qu'un auditeur ne perçoit pas la musique comme une simple séquence de sons, pas plus que le compositeur n'a conçu son morceau comme tel. La musique est en effet constituée de motifs dont l'organisation intrinsèque et les relations mutuelles participent à la structuration du propos musical, et ce à plusieurs échelles simultanément. Cependant, il est aujourd'hui encore très difficile de définir précisément le terme de concept musicale. L'un des principaux aspects de la musique est qu'elle est en grande partie constituée de redondances, sous forme de répétitions exactes et variées. L'organisation de ces redondances permet de susciter une attente chez l'auditeur. Une surprise peut alors être créée en présentant des éléments qui ne correspondent pas à cette attente. Ce travail de thèse se base sur l'hypothèse que les redondances, l'attente et la surprise sont des éléments essentiels pour la description de la structure musicale d'un segment. Un certain nombre de questions découlent de ce constat: quels sont les éléments musicaux qui participent à la structure d'un objet musical ? Quelles sont les dépendances entre ces éléments qui jouent un rôle essentiel dans la structuration d'un objet musical ? Comment peut-on décrire une relation entre deux éléments musicaux tels que des accords, des motifs rythmiques ou mélodiques ? Dans ce manuscrit, des éléments de réponse sont proposés par la formalisation et l'implémentation d'un modèle multi-échelle de description de la structure d'un segment musical : les Graphes Polytopiques à Relations Latentes (GPRL/PGLR). Dans ce travail, les segments considérés sont les sections successives qui forment une pièce musicale. Dans le cas de la pop, genre musical sur lequel se concentre ce travail, il s'agit typiquement d'un couplet ou d'un refrain, de 15 sec. environ, comprenant un début et une fin bien définis. En suivant le formalisme PGLR, les relations de dépendance prédominantes entre éléments musicaux d'un segment sont celles qui relient les éléments situés à des positions homologues sur la grille métrique du segment. Cette approche généralise sur le plan multi-échelle le modèle Système&Contraste qui décrit sous la forme d'une matrice 2×2 le système d'attente logique au sein d'un segment et la surprise qui découle de la réalisation de cette attente. Pour des segments réguliers de taille 2^n, le PGLR peut être représenté sur un n-cube (carré, cube, tesseract, ...), où n est le nombre d'échelles considérées. Chaque nœud du polytope correspond à un élément musical fondamental (accord, motif, note...), chaque arête représente une relation entre deux nœuds et chaque face représente un système de relations. La recherche du PGLR correspondant à la meilleure description de la structure d'un segment musical s'opère par l'estimation jointe : de la description du polytope (un n-polytope plus ou moins régulier) ; de la configuration du graphe sur le polytope, permettant de décrire le flux de dépendance et les interactions entre les éléments par des implications élémentaires systémiques au sein du segment ; la description de l'ensemble des relations entre les nœuds du graphe. Le but du modèle PGLR est à la fois de décrire les dépendances temporelles entre les éléments d'un segment et de modéliser l'attente logique et la surprise qui découlent de l'observation et de la perception des similarités et des différences entre ces éléments. Cette approche a été formalisée et implémentée pour décrire la structure de séquences d'accords ainsi que de segments rythmiques et mélodiques, puis évaluée par sa capacité à prédire des segments inconnus. La mesure utilisée pour cette évaluation est la perplexité croisée calculée à partir des données du corpus RWC POP. Les résultats obtenus donnent un large avantage à la méthode multi-échelle proposée, qui semble mieux à même de décrire efficacement la structure des segments testés
In this thesis, we approach these questions by defining and implementing a multi-scale model for music segment structure description, called Polytopic Graph of Latent Relations (PGLR). In our work, a segment is the macroscopic constituent of the global piece. In pop songs, which is the main focus here, segments usually correspond to a chorus or a verse, lasting approximately 15 seconds and exhibiting a clear beginning and end. Under the PGLR scheme, relationships between musical elements within a musical segment are assumed to be developing predominantly between homologous elements within the metrical grid at different scales simultaneously. This approach generalises to the multi-scale case the System&Contrast framework which aims at describing, as a 2×2 square matrix, the logical system of expectation within a segment and the surprise resulting from that expectation. For regular segments of 2^n events, the PGLR lives on a n-dimensional cube (square, cube, tesseract, etc...), n being the number of scales considered simultaneously in the multi-scale model. Each vertex in the polytope corresponds to a low-scale musical element, each edge represents a relationship between two vertices and each face forms an elementary system of relationships. The estimation of the PGLR structure of a musical segment can then be obtained computationally as the joint estimation of : the description of the polytope (as a more or less regular n-polytope) ; the nesting configuration of the graph over the polytope, reflecting the flow of dependencies and interactions as elementary implication systems within the musical segment, the set of relations between the nodes of the graph. The aim of the PGLR model is to both describe the time dependencies between the elements of a segment and model the logical expectation and surprise that can be built on the observation and perception of the similarities and differences between elements with strong relationships. The approach is presented conceptually and algorithmically, together with an extensive evaluation of the ability of different models to predict unseen data, measured using the cross-perplexity value. These experiments have been conducted both on chords sequences, rhythmic and melodic segments extracted from the RWC POP corpus. Our results illustrate the efficiency of the proposed model in capturing structural information within such data