Thèses sur le sujet « Medical image sequences »

This thesis consider enhancement of X-ray fluoroscopy image sequences. The purpose is to investigate the possibilities to improve the image enhancement in Biplanar 500, a fluoroscopy system developed by Swemac Medical Appliances, for use in orthopedic surgery.

An algorithm based on recursive filtering, for temporal noise suppression, and motion compensation, for avoidance of motion artifacts, is developed and tested on image sequences from the system. The motion compensation is done both globally, by using the theory of the shift theorem, and locally, by subtracting consecutive frames. Also a new type of contrast adjustment is presented, received with an unlinear mapping function.

The result is a noise reduced image sequence that shows no blurring effects upon motion. A brief study of the result shows, that both the image sequences with this algorithm applied and the contrast adjusted images are preferred by orthopedists compared to the present images in the system.

The aim of this work is to propose an algorithm to segment and track the myocardium using the level-set formalism. The myocardium is first approximated by a geometric model (hyperquadrics) which allows to handle asymetric shapes such as the myocardium while avoiding a learning step. This representation is then embedded into the level-set formalism as a shape prior for the joint segmentation of the endocardial and epicardial borders. This shape prior term is coupled with a local data attachment term and a thickness term that prevents both contours from merging. The algorithm is validated on a dataset of 80 images at end diastolic and end systolic phase with manual references from 3 cardiologists. In a second step, we propose to segment whole sequences using motion information. To this end, we apply a level conservation constraint on the implicit function associated to the level-set and express this contraint as an energy term in a variational framework. This energy is then added to the previously described algorithm in order to constrain the temporal evolution of the contour. Finally the algorithm is validated on 20 echocardiographic sequences with manual references of 2 experts (corresponding to approximately 1200 images).

Radiotherapy plays an increasingly important role in cancer treatment, and medical imaging plays an increasingly important role in radiotherapy. Magnetic resonance imaging (MRI) is poised to be a major component in the development towards more effective radiotherapy treatments with fewer side effects. This thesis attempts to contribute in realizing this potential. Radiotherapy planning requires simulation of radiation transport. The necessary physical properties are typically derived from CT images, but in some cases only MR images are available. In such a case, a crude but common approach is to approximate all tissue properties as equivalent to those of water. In this thesis we propose two methods to improve upon this approximation. The first uses a machine learning algorithm to automatically identify bone tissue in MR. The second, which we refer to as atlas-based regression, can be used to generate a realistic, patient-specific, pseudo-CT directly from anatomical MR images. Atlas-based regression uses deformable registration to estimate a pseudo-CT of a new patient based on a database of aligned MR and CT pairs. Cancerous tissue has a dierent structure from normal tissue. This affects molecular diusion, which can be measured using MRI. The prototypical diusion encoding sequence has recently been challenged with the introduction of more general  waveforms. To take full advantage of their capabilities it is, however, imperative to respect the constraints imposed by the hardware while at the same time maximizing the diffusion encoding strength. In this thesis we formulate this as a constrained optimization problem that is easily adaptable to various hardware constraints.

Les hôpitaux et les centres médicaux produisent une énorme quantité d'images médicales numériques chaque jour, notamment sous la forme de séquences d'images. En raison de la grande capacité de stockage et de la bande passante de transmission limitée, une technique de compression efficace est nécessaire. Nous avons proposé un algorithme de compression de séquences d'images médicales MMWaaves. Il repose sur l'utilisation de modèles Markovien couplé avec le codeur Waaves de la société Cira qui est certifié en tant que dispositif médicale. Nous avons démontré que MMWaaves a apporté un gain de compression supérieur à 30% par rapport à JPEG2000 et Waaves tout en gardant la qualité nécessaire pour les diagnostics cliniques (SSIM>0.98). En outre, il a permis d'atteindre des taux de compression égaux à ceux obtenus par H.264 en améliorant la qualité. Ensuite, nous avons développé une nouvelle chaine de compression MLPWaaves à base de différence en DWT suivie d'un nouveau modèle de tri adaptatif LPEAM permettant l'optimisation de la stationnarité locale des coefficients. Nous avons obtenu un gain de compression allant à 80% par rapport à Waaves et JPEG2000 tout en assurant une qualité exceptionnelle pour le diagnostic médical. Finalement, afin de transmettre à distance les images médicales du centre de santé à l'appareil mobile du médecin, nous avons proposé un système de télé-radiologie pour le codage et le décodage basé sur nouveau paradigme multithreading. La validation de cette nouvelle solution a été réalisée sur deux plateformes différentes. Nous avons obtenu un facteur d'accélération égal à 5 sur un Intel Core i7-2600 et un facteur égal à 3 sur une tablette Samsung Galaxy
Hospitals and medical centers produce an enormous amount of digital medical images every day especially in the form of image sequences. Due to the large storage size and limited transmission and width, an efficient compression technique is necessary. We first proposed a compressor algorithm for medical images sequences MMWaaves. It is based on Markov fields coupled with the certified medical device Waaves of Cira company. We demonstrated that MMWaaves provided a compression gains greater than 30% compared to JPEG2000 and Waaves while ensuring outstanding image quality for medical diagnosis (SSIM> 0.98). In addition, it achieved compression rates equal to those obtained by H.264 while improving the image quality. Then we developed a new compression algorithm MLPWaaves based on DWT difference followed by a new adaptive scanning model LPEAM in order to optimize the local stationary of wavelet coefficients. We obtained a compression gain up to 80% compared to Waaves and JPEG2000 while ensuring exceptional quality for medical diagnosis. Finally, in order to transmit medical images for diagnostic from the health center to the mobile device of the doctor, we proposed client-server remote radiology system for encoding and decoding. It is based on a multithreading paradigm to accelerate treatment. The validation of this solution was performed on two different platforms. We achieved an acceleration factor of 5 on an Intel Core i7-2600 and a factor of 3 on Samsung Galaxy tablet

Les malalties cardiovasculars són una de les principals causes de mortalitat als països desenvolupats. La majoria d'elles són degudes a malalties arterials (especialment les coron ries), que vénen causades per l'acumulació de placa. Aquesta patologia estreny el flux sanguini (estenosi) i afecta les propietats elàstiques i bio-mecàniques (arteriosclerosi) de les artèries. En les últimes dècades, l'Ecografia Intra-Coronària (EIC) ha esdevingut una tècnica usual de diagnòstic per la imatge i seguiment de les malalties coronàries. L'EIC està basada en un cateterisme que mostra una seqüència d'imatges corresponents a seccions de l'artèria sota estudi. La inspecció visual de cadascuna d'aquestes imatges proporciona informació sobre el percentatge d'estenosi, mentre que la inspecció de les vistes longitudinals propociona informació sobre les propietats bio-mecàniques, que pot prevenir un desenllaç fatal de la malaltia cardiovascular. Per una banda, la dinàmica arterial (deguda al batec del cor, entre d'altres) és un dels principals artefactes per poder explorar les propietats biomecàniques. Al mateix temps, les mesures manuals d'estenosi requereixen un traçat manual de les vores del vas, tasca feixuga que consumeix molt de temps i que pot patir variabilitat entre observadors.
Aquesta tesi proposa vàries eines de processament d'imatge per explorar la dinàmica de les artèries i les seves estructures. Presentem un model físic per extreure, analitzar i corregir la dinàmica rígida transversal dels vasos i per recuperar la fase cardíaca. A més, introduïm un mètode estadístic-determinista per a la detecció automàtica de les vores del vas. En particular, l'enfoquem a segmentar l'adventícia. Un protocol de validació acurat per assegurar una aplicació clínica fiable dels mètodes és un pas crucial en qualsevol proposta d'algorisme. En aquesta tesi tenim especial cura de dissenyar protocols de validació per a cadascuna de les tècniques proposades i contribuïmm a la validació de la dinàmica in vivo amb un indicador objectiu i quantitatiu per mesurar la quantitat de moviment suprimida.
Cardiovascular diseases are a leading cause of death in developed countries. Most of them are caused by arterial (specially coronary) diseases, mainly caused by plaque accumulation. Such pathology narrows blood flow (stenosis) and affects artery bio-mechanical elastic properties (atherosclerosis). In the last decades, IntraVascular UltraSound (IVUS) has become a usual imaging technique for the diagnosis and follow up of arterial diseases. IVUS is a catheter-based imaging technique which shows a sequence of cross sections of the artery under study. Inspection of a single image gives information about the percentage of stenosis. Meanwhile, inspection of longitudinal views provides information about artery bio-mechanical properties, which can prevent a fatal outcome of the cardiovascular disease. On one hand, dynamics of arteries (due to heart pumping among others) is a major artifact for exploring tissue bio-mechanical properties. On the other one, manual stenosis measurements require a manual tracing of vessel borders, which is a time-consuming task and might suffer from inter-observer variations.
This PhD thesis proposes several image processing tools for exploring vessel dynamics and structures. We present a physics-based model to extract, analyze and correct vessel in-plane rigid dynamics and to retrieve cardiac phase. Furthermore, we introduce a deterministic-statistical method for automatic vessel borders detection. In particular, we address adventitia layer segmentation. An accurate validation protocol to ensure reliable clinical applicability of the methods is a crucial step in any proposal of an algorithm. In this thesis we take special care in designing a validation protocol for each approach proposed and we contribute to the in vivo dynamics validation with a quantitative and objective score to measure the amount of motion suppressed.

Corneal diseases are one of the major causes of visual impairment and blindness worldwide. Used for diagnoses, a laser confocal microscope provides a sequence of images, at incremental depths, of the various corneal layers and structures. From these, ophthalmologists can extract clinical information on the state of health of a patient’s cornea. However, many factors impede ophthalmologists in forming diagnoses starting with the large number and variable quality of the individual images (blurring, non-uniform illumination within images, variable illumination between images and noise), and there are also difficulties posed for automatic processing caused by eye movements in both lateral and axial directions during the scanning process. Aiding ophthalmologists working with long sequences of corneal image requires the development of new algorithms which enhance, correctly order and register the corneal images within a sequence. The novel algorithms devised for this purpose and presented in this thesis are divided into four main categories. The first is enhancement to reduce the problems within individual images. The second is automatic image classification to identify which part of the cornea each image belongs to, when they may not be in the correct sequence. The third is automatic reordering of the images to place the images in the right sequence. The fourth is automatic registration of the images with each other. A flexible application called CORNEASYS has been developed and implemented using MATLAB and the C language to provide and run all the algorithms and methods presented in this thesis. CORNEASYS offers users a collection of all the proposed approaches and algorithms in this thesis in one platform package. CORNEASYS also provides a facility to help the research team and Ophthalmologists, who are in discussions to determine future system requirements which meet clinicians’ needs.
The data and image files accompanying this thesis are not available online.

Corneal diseases are one of the major causes of visual impairment and blindness worldwide. Used for diagnoses, a laser confocal microscope provides a sequence of images, at incremental depths, of the various corneal layers and structures. From these, ophthalmologists can extract clinical information on the state of health of a patient’s cornea. However, many factors impede ophthalmologists in forming diagnoses starting with the large number and variable quality of the individual images (blurring, non-uniform illumination within images, variable illumination between images and noise), and there are also difficulties posed for automatic processing caused by eye movements in both lateral and axial directions during the scanning process. Aiding ophthalmologists working with long sequences of corneal image requires the development of new algorithms which enhance, correctly order and register the corneal images within a sequence. The novel algorithms devised for this purpose and presented in this thesis are divided into four main categories. The first is enhancement to reduce the problems within individual images. The second is automatic image classification to identify which part of the cornea each image belongs to, when they may not be in the correct sequence. The third is automatic reordering of the images to place the images in the right sequence. The fourth is automatic registration of the images with each other. A flexible application called CORNEASYS has been developed and implemented using MATLAB and the C language to provide and run all the algorithms and methods presented in this thesis. CORNEASYS offers users a collection of all the proposed approaches and algorithms in this thesis in one platform package. CORNEASYS also provides a facility to help the research team and Ophthalmologists, who are in discussions to determine future system requirements which meet clinicians’ needs.

L'objet de cette these est de proposer et de developper une methode generale et robuste permettant de traiter deux types de problematique de traitement de sequences. L'approche generale se fonde sur une modelisation pertinente du phenomene observe et sur l'inference statistique des parametres de cette modelisation et de leurs variations. Pour les deux problematiques, nous avons developpe des systemes hierarchiques et hybrides, tirant parti des capacites d'approximateurs universels des reseaux de neurones et integrant de la connaissance a priori sur le probleme. Le premier probleme traite s'inscrit dans le cadre general de la prevision du comportement d'un systeme dynamique evoluant selon le contexte. Plus precisement, nous avons propose un nouveau systeme hierarchique et evolutif pour prevoir la fin de decharge de batteries rechargeables alimentant un appareil portable. Le modele original propose utilise deux reseaux de neurone. Le premier est un simple modele d'une courbe de decharge alors que le second est responsable de l'adaptation aux donnees contextuelles et estime les parametres du premier reseau. La version incrementale proposee permet une adaptation en ligne aux variabilites comportementales des batteries. Les resultats obtenus sont bons avec une erreur moyenne de 6 minutes pour un evenement qui peut intervenir dans un intervalle de 10 heures. Le second probleme aborde consiste en l'extraction automatique d'un contour dans des images medicales : le contour du ventricule gauche dans des images angiocardiographiques. La delineation precise de cet objet sert de base a la mesure de quantites tres utiles au diagnostic de maladies cardio-vasculaires. Le systeme propose utilise le plus possible de l'information a priori de haut niveau afin de restreindre la recherche vers le contour le plus probable. Cette recherche se base sur un modele hybride reseau de neurones - chaine de markov cachee. Les resultats prometteurs obtenus demontrent l'interet de cette demarche.

Multi-spectral images have the advantage of providing complementary information to resolve some ambiguities. But, the challenge is how to make use of the multi-spectral images effectively. In this thesis, our study focuses on the fusion of multi-spectral images by extracting the most useful features to obtain the best segmentation with the least cost in time. The Support Vector Machine (SVM) classification integrated with a selection of the features in a kernel space is proposed. The selection criterion is defined by the kernel class separability. Based on this SVM classification, a framework to follow up brain tumor evolution is proposed, which consists of the following steps: to learn the brain tumors and select the features from the first magnetic resonance imaging (MRI) examination of the patients; to automatically segment the tumor in new data using a multi-kernel SVM based classification; to refine the tumor contour by a region growing technique; and to possibly carry out an adaptive training. The proposed system was tested on 13 patients with 24 examinations, including 72 MRI sequences and 1728 images. Compared with the manual traces of the doctors as the ground truth, the average classification accuracy reaches 98.9%. The system utilizes several novel feature selection methods to test the integration of feature selection and SVM classifiers. Also compared with the traditional SVM, Fuzzy C-means, the neural network and an improved level set method, the segmentation results and quantitative data analysis demonstrate the effectiveness of our proposed system.

L’objectif principal de l’imagerie par résonance magnétique (IRM) est d’apporter des renseignements pour le diagnostic clinique. A partir de séquences d’imagerie agissant sur le comportement des aimantations microscopiques, il est possible d’avoir accès à une source précieuse d’information macroscopique. Dans cette thèse, nous étudions un type de séquence adapté à l’examen de tissus durs comme par exemple le cartilage. Ces séquences présentent l’avantage de moduler l’interaction dipolaire présente dans les tissus. La séquence de contraste dipolaire, qui a servi de base à nos travaux est dérivée d’une séquence appelée Sandwich d’écho magique (MSE), qui permet de modifier l’interaction dipolaire. Initialement développée pour sonder des matériaux extrêmement durs, elle avait été modifiée au laboratoire pour être utilisée sur des tissus biologiques « moins solides ». Elle permettait également de gagner deux ordres de grandeur en temps d’acquisition, ce qui la rendait compatible avec un contexte clinique. Le principal but de ce travail de thèse est de préciser les contextes de mise en oeuvre de cette séquence et de la comparer à d’autres types de séquences (écho de spin, écho stimulé et élastographie) pour en déduire de nouveaux paramètres d’intérêt. Nous avons travaillé sur des échantillons qui ont des propriétés proches des matériaux solides : des polymères de type plastisol®. Cette étude est un support de réflexion sur les cadres d’application des séquences dipolaires de type MSE pour le diagnostic
The main objective of magnetic resonance imaging is to provide information for clinical diagnosis. From imaging sequences acting on the behaviour of microscopic magnetisation, it is possible to have access to a valuable source of macroscopic information. In this thesis, we study a sequence type adapted to the investigation of rigid tissues such as cartilage. The principal advantage of these sequences is to modulate the dipolar interaction present in the tissues. The dipolar contrast sequence, which served as a basis for our work, is derived from a RMN sequence called Magic Sandwich Echo (MSE) that allows us to modify the dipolar interaction. Initially developed to probe extremely solid materials, it had been modified in the laboratory to be used on "less solid" biological tissues. With this version the acquisition time has been reduced by two orders of magnitude, which makes this method compatible with clinical context. The original purpose of this thesis work is to specify the context of implementation of this sequence and to compare it to other types of sequences (spin echo, stimulated echo and elastography) to deduce new parameters of interest. We have worked on samples that have properties close to solid materials: polymers of plastisol® type. Therefore this study gives the application framework of dipolar sequences of the MSE type for diagnosis

L’analyse d’images médicales joue un rôle essentiel en cardiologie pour la réalisation du diagnostique cardiaque clinique et le suivi de l’état du patient. Parmi les modalités d’imagerie utilisées, l’imagerie par ultrasons, temps réelle, moins coûteuse et portable au chevet du patient, est de nos jours la plus courante. Malheureusement, l’étape nécessaire de segmentation sémantique (soit l’identification et la délimitation précise) des structures cardiaques est difficile en échocardiographie à cause de la faible qualité des images ultrasonores, caractérisées en particulier par l’absence d’interfaces nettes entre les différents tissus. Pour combler le manque d’information, les méthodes les plus performante, avant ces travaux, reposaient sur l’intégration d’informations a priori sur la forme ou le mouvement du cœur, ce qui en échange réduisait leur adaptabilité au cas par cas. De plus, de telles approches nécessitent pour être efficaces l’identification manuelle de plusieurs repères dans l’image, ce qui rend le processus de segmentation difficilement reproductible. Dans cette thèse, nous proposons plusieurs algorithmes originaux et entièrement automatiques pour la segmentation sémantique d’images échocardiographiques. Ces méthodes génériques sont adaptées à la segmentation échocardiographique par apprentissage supervisé, c’est-à-dire que la résolution du problème est construite automatiquement à partir de données pré- analysées par des cardiologues entraînés. Grâce au développement d’une base de données et d’une plateforme d’évaluation dédiées au projet, nous montrons le fort potentiel clinique des méthodes automatiques d’apprentissage supervisé, et en particulier d’apprentissage profond, ainsi que la possibilité d’améliorer leur robustesse en intégrant une étape de détection automatique des régions d’intérêt dans l’image
The analysis of medical images plays a critical role in cardiology. Ultrasound imaging, as a real-time, low cost and bed side applicable modality, is nowadays the most commonly used image modality to monitor patient status and perform clinical cardiac diagnosis. However, the semantic segmentation (i.e the accurate delineation and identification) of heart structures is a difficult task due to the low quality of ultrasound images, characterized in particular by the lack of clear boundaries. To compensate for missing information, the best performing methods before this thesis relied on the integration of prior information on cardiac shape or motion, which in turns reduced the adaptability of the corresponding methods. Furthermore, such approaches require man- ual identifications of key points to be adapted to a given image, which makes the full process difficult to reproduce. In this thesis, we propose several original fully-automatic algorithms for the semantic segmentation of echocardiographic images based on supervised learning ap- proaches, where the resolution of the problem is automatically set up using data previously analyzed by trained cardiologists. From the design of a dedicated dataset and evaluation platform, we prove in this project the clinical applicability of fully-automatic supervised learning methods, in particular deep learning methods, as well as the possibility to improve the robustness by incorporating in the full process the prior automatic detection of regions of interest

Ce travail a pour objectif de contribuer a l'identification des resonances obtenues in vivo par spectroscopie rmn proton 1d et 2d, globale et localisee. Les techniques d'edition spectrale et de caracterisation developpees ont ete appliquees en particulier a l'etude des tumeurs intracerebrales du cerveau de rat in vivo. Un travail preliminaire a consiste a elaborer une methode rapide de correction des imperfections introduites par un modulateur de phase et d'amplitude, qui permet d'obtenir des impulsions de bonne qualite. Plusieurs axes de recherche ont ensuite ete suivis afin d'apporter des informations complementaires contribuant a l'identification moleculaire. Un de ces axes a concerne les techniques de rmn a deux dimensions, en examinant plus particulierement les fonctions d'apodisation utilisees pour le traitement des spectres 2d obtenus in vivo, afin d'optimiser le rapport signal sur bruit. La spectroscopie 2d de correlation combinee a l'imagerie spectroscopique a permis d'acceder a la repartition spatiale des acides amines et des carbohydrates presents dans une plante in vivo. D'autre part, les techniques d'observation localisees (imagerie et imagerie spectroscopique) des coherences a multiples quanta ont ete proposees afin d'observer simultanement les differents ordres de coherence des systemes de spins couples et non-couples. Enfin, dans un autre contexte, la mesure localisee du coefficient de diffusion a conduit a l'edition et a la caracterisation des metabolites presents dans les tumeurs intracerebrales in vivo en fonction de leur degre de mobilite translationnelle

博士
國立成功大學
資訊工程學系碩博士班
97
Motion analysis is very useful for recognizing target patterns from a sequence of images. Applications in motion estimation and target tracking become especially important in medical and biomedical researches nowadays. However, traditional methods which are optimal for rigid body motion are not suitable for medical analysis due to the object deformation and noise problems. In this study, we tried to propose adequate motion estimation methods for several medical motion applications which include motion field estimation from ultrasound images, tag line tracking from tagged magnetic resonance (MR) images, and live cell tracking from microscopic images. Generally, the usual problems in medical motion analysis include: speckle noises and temporal de-correlation of the speckle patterns in ultrasound images; large motion and tag decaying problems in tagged MR images; and low contrast in pseudopods and topological changes in cellular microscopic images. To overcome these problems, it is necessary to integrate a priori knowledge based on the physical properties into the motion estimation process. In this study, we first designed a hierarchical maximum a posteriori estimator together with an ultrasonic feature model for ultrasound image sequences. A motion compounding method is also proposed to reduce speckle noises and to enhance image quality based on the proposed motion estimation method. To cope with the problems of large motion and tag decaying, we proposed to incorporate a cardiac motion model based prediction scheme and a candidate pre-screening technique together with the deformable models to track the tag lines. To segment and track highly deformable cells, we have presented an automatic method based on the framework of modified T-snakes coupled with the knowledge of cellular life model. The proposed motion estimation methods were compared with several existing methods via a series of experiments with both simulated and clinical image sequences. Experimental results showed that motion could be accurately assessed in different types of imaging modalities. The proposed systems can help to perform better quantification and analyses in clinical applications. It will certainly help medical doctors to achieve better observation and more accurate assessments, and thus result in better diagnostic quality.

博士
中原大學
電子工程研究所
95
The enormous data of medical image sequences bring a transmission and storage problem that can be solved by using a compression technique. For the lossy compression of a very long medical image sequence, automatically maintaining the diagnosis features in reconstructed images is essential. In this dissertation, the proposed wavelet-based adaptive vector quantizer incorporates a distortion-constrained codebook replenishment (DCCR) mechanism to meet a user-defined quality demand in peak signal-to-noise ratio (PSNR). Combining a codebook updating strategy and the well-known SPIHT technique, the DCCR mechanism provides an excellent coding gain. Experimental results show that the proposed approach is superior to the pure SPIHT and the JPEG2000 algorithms in terms of coding performance. We also propose an iterative fast searching algorithm to find the desired signal quality along an energy-quality curve instead of a traditional rate-distortion curve. The algorithm performs the quality control quickly, smoothly, and reliably. Due to the disadvantages of using codebooks (high memory consumption, time-consuming initial codebook training, and serious error propagation) and the need of region-based coding with lossy-to-lossless functionality for medical image compression, we further propose a region-based compression approach for medical image sequences in this dissertation. To keep comparable coding performance with the DCCR mechanism, the new approach replaces codebook searching and DCCR by a motion estimation and compensation mechanism in wavelet coefficient domain to avoid the aforementioned disadvantages of using codebooks. Furthermore, the lossy and/or lossless quality levels of multiple regions in a decompressed image can be specified by a physician and achieved automatically by the proposed approach. After performing the shape-adaptive discrete wavelet transform (DWT), DWT coefficients arranged in hierarchical trees are grouped into wavelet blocks (WBs). The multiple regions with arbitrary shapes can be specified to have different assigned quality levels in terms of PSNR. The quality demand in pixel domain is converted to the equivalent distortion allowed for each WB in DWT domain via the use of DWT properties and a distortion assignment strategy. Experimental results show that the proposed approach has much smaller quality variation than the conventional SPIHT coding method. For image sequences with specific regions in each image, the proposed method achieves 40 and 48 dB target PSNR subject to the maximum quality variation of only 7% and 1%, respectively. For still images with arbitrarily shaped objects and various modalities, the quality control of the proposed method is still very effective. In addition, the performance of controlling lossy and/or lossless quality levels of multiple regions in one image is demonstrated and the quality controlling accuracy using SA-DWT with different filter banks is compared. Finally, the proposed approach not only performs better than SPIHT in coding performance, but also preserves the boundary and shape of the specified region accurately.

碩士
國立屏東商業技術學院
資訊工程系(所)
102
Bio-inspired computing has been widely used in various fields. It simulates animal or insect behavior, such as clustering or procreation acts, to design the algorithms for solving optimization problems. In recent years there are a variety of bio-inspired computing algorithms have been proposed, such as firefly algorithm, artificial bee colony algorithm. In this paper, we use firefly algorithm to design the motion estimation, based on block-matching technology. The algorithm is applied to the extensor tendon and achilles tendon ultrasonic image sequence and is proven its effectiveness compared with the traditional full search.