Auswahl der wissenschaftlichen Literatur zum Thema „Identity-utility extraction“

Background. Published utility estimates for lung cancer are plentiful and vary greatly. The reason for this variability is unclear, but may result from differences in the methods used to elicit each utility. Purpose. To identify a set of pooled lung cancer utility estimates reflective of the available literature and determine which methodological factors significantly influence the value of lung cancer utility. Data Sources. Searches of PubMed, the NHS Economic Evaluation Database, and the Cost Effectiveness Analysis Registry from the Center for the Evaluation of Value and Risk in Health. Study Selection. English-language studies were included if they presented at least one previously unpublished lung cancer utility value, noted the elicitation technique and utility value provider. Data Extraction and Analysis. Two trained readers independently reviewed each article and extracted information for analysis. A hierarchical linear model (HLM) was used to perform a meta-regression with cancer stage, lower bound of scale, upper bound of scale, respondent, elicitation method, and lung cancer subtype as explanatory variables. Data Synthesis. Twenty-three articles containing 223 unique utility values were included. Lung cancer stage and subtype, the upper bound label of the utility scale, and respondent identity were significant predictors of utility (P < 0.05), while the lower bound label of utility scale was not. The HLM provided a set of pooled utility values for metastatic (0.57), mixed or nonspecified stage (0.77), and nonmetastatic lung cancer (0.87 )—for the case of standard gamble as method, patients as respondents, non-small-cell lung cancer and scale labeled death to perfect health. Conclusion. Methodological factors significantly affect lung cancer utilities; therefore, analysts should avoid direct comparisons of lung cancer utility values elicited with dissimilar methods.

A 2-y-old Brahman bull was presented with progressive hindlimb ataxia and paraparesis that led to recumbency. Postmortem examination revealed scattered pinpoint, red-brown foci within the brainstem and gray matter of the spinal cord, and a larger lesion within the spinal cord at the level of T13. Histology of the section of T13 contained cross-sections of nematodes consistent with Parelaphostrongylus tenuis. Evidence of inflammation was present in other affected areas of the spinal cord and brain. DNA extraction and nested PCR were performed, which demonstrated 98% identity and 100% coverage to both P. tenuis and P. andersoni. Our case highlights the utility of DNA sequencing in parasite identification.

As a weak high-frequency signal embedded in digital images, Photo Response Non-Uniformity (PRNU) is particularly vulnerable to interference from low-frequency components during the extraction process, which affects its reliability in real-world forensic applications. Previous studies have not successfully identified the effective frequency band of PRNU, leaving low-frequency interference insufficiently suppressed and impacting PRNU’s utility in scenarios such as source camera identification, image integrity verification, and identity verification. Additionally, due to differing operational mechanisms, current mainstream PRNU enhancement algorithms cannot be integrated to improve their performance further. To address these issues, we conducted a frequency-by-frequency analysis of the estimated PRNU and discovered that it predominantly resides in the frequency band above 10 Hz. Based on this finding, we propose a guided-filtering PRNU enhancement algorithm. This algorithm can function as a plug-and-play module, seamlessly integrating with existing mainstream enhancement techniques to further boost PRNU performance. Specifically, we use the PRNU components below 10 Hz as a guide image and apply guided filtering to reconstruct the low-frequency interference components. By filtering out these low-frequency components, we retain and enhance the high-frequency PRNU signal. By setting appropriate enhancement coefficients, the low-frequency interference is suppressed, and the high-frequency components are further amplified. Extensive experiments on publicly available Dresden and Daxing digital device forensics datasets confirm the efficiency and robustness of the proposed method, making it highly suitable for reliable forensic analysis in practical settings.

Introduction: Isolation of fungi from tissue specimens using conventional methods is time consuming. However, in some cases, the histopathological examination (HPE) of tissue alone is unable to provide a definite identity of the fungus. Alternatively, a non-culture method, such as polymerase chain reaction (PCR) detecting the internal transcribed spacer (ITS) rRNA genes of the fungi, is a promising diagnostic tool for rapid and accurate diagnosis of tissue mycoses. Methods: This work investigated the utility of panfungal PCR in identifying agents of tissue mycoses in 87 FFPE archive specimens. Deoxyribonucleic acid (DNA) extraction was performed on FFPE specimens by using QIAamp DNA FFPE Tissue Kit. The ITS2 region was amplified using ITS3/ITS4 primers. The PCR products were sequenced using the same primers and compared to the NCBI nucleotide database for species identification. Results: Fungal DNA was successfully amplified in 52 (59.8%) specimens, from which only 23 (44.0%) fungi were consistent with clinical/HPE findings. The identified fungi were Aspergillus spp., Candida spp., Penicillium spp., Cryptococcus neoformans, Talaromyces marneffei, and Rhizopus oryzae. A few rare fungi were also identified, such as Diaporthe longicolla and fungus-like oomycete such as Pythium insidiosum that are commonly associated with plant pathogens. Conclusion: Although PCR was able to offer accurate genus/species identification, utilising this method on paraffinised tissue specimens must be evaluated by considering many factors that will reduce its sensitivity and specificity. Therefore, it is important to correlate the PCR results with clinical and HPE findings to obtain a correct diagnosis and adequate treatment for tissue mycoses.

Abstract Introduction: Pancreatic ductal adenocarcinoma (PDAC) remains challenging to treat in part due to the dense desmoplastic stroma characteristic of the tumor microenvironment (TME). Often, less than 50% of the tumor mass is composed of malignant ductal epithelial cells. The relationship between the TME and the ductal tumor is important in tumorigenesis, metastasis, and treatment resistance. Cancer associated fibroblasts (CAFs) are an abundant cell type in the TME and have a dynamic role in both promoting and inhibiting tumor progression. Efforts to target fibroblast function therapeutically have yielded mixed results. While pre-clinical data have generated some optimism that CAF inhibition may improve clinical outcomes, this has not translated to the clinical setting. One likely reason for these inconsistent findings is the evolving understanding of CAF heterogeneity. Expanding our understanding of the interactions between the epithelial tumor and CAFs may identify new modalities for modulating the TME to augment treatment efficacy and improve patient outcomes. This requires the generation of human model systems to better interrogate intercellular interactions. Methods: We have developed an ex vivo model system to investigate interactions between ductal epithelial tumor cells and CAFs using matched patient-derived organoids (PDOs) and CAFs extracted following dissociation of surgical specimens. Flow cytometry was used to evaluate the viability of both cellular components as well as associated proliferation and cell identity. Moreover, we have used immunohistochemistry to further characterize this system while preserving the spatial relationship of the cell populations. Most recently, we employed the MULTI-seq multiplex single cell RNA sequencing (scRNA-seq) pipeline to explore the transcriptional dynamics of this system at intervals over 96 hours by examining 4 conditions with 1,500 cells sequenced per condition per time point. Results: Flow cytometry demonstrated that both PDOs and CAFs remain viable in this co-culture construct and that both cell types can be reliably identified using epithelial markers to identify organoid cells and an exclusion panel to identify the CAFs. We have further validated these findings using immunohistochemistry and demonstrated that in co-culture both PDOs and CAFs remain viable and continue to proliferate. We have also demonstrated successful extraction of both cell types from the Matrigel scaffold in which they are grown for use in downstream high dimensional genomics assays. Conclusions: We have previously demonstrated the utility of a PDO model to explore clinically relevant biological mechanisms in PDAC. However, the addition of CAFs to this model provides a more comprehensive representation of the in vivo tumor. This ex vivo system is viable and flexible for multiple end point assays and can be used to explore relevant cellular interactions and mechanisms. Our ongoing investigation will specifically focus on PDO-CAF interactions and transcriptional dynamics over time as we interrogate the scRNA-seq data. Citation Format: Jacquelyn W. Zimmerman, Genevieve Stein-O'Brien, Richard A. Burkhart, Elana J. Fertig, Elizabeth M. Jaffee. Patient-derived organoids and cancer associated fibroblasts as a co-culture model to explore cell type interactions in pancreatic cancer [abstract]. In: Proceedings of the AACR Virtual Special Conference on Pancreatic Cancer; 2021 Sep 29-30. Philadelphia (PA): AACR; Cancer Res 2021;81(22 Suppl):Abstract nr PO-080.

We previously demonstrated that erythroblastic island macrophages (Mφs) in the mouse bone marrow (BM) support not only differentiating erythroblasts but also granulocyte precursors, proposing the term erythromyeloblastic islands (EMBIs) for these niches of terminal hematopoiesis (Romano, Seu, et al. Blood 2022). However, the identity, function, and regulation of the EMBI Μφs remains poorly understood due to their immunophenotypic heterogeneity and many technical hurdles such as strong autofluorescence, tight interactions with associated cells, and susceptibility to fragmentation with mechanical disruption; indeed, fragmentation of EMBI Μφs has been recently demonstrated by our team and others to confound immunophenotyping and transcriptomic studies. Our previous single cell RNA-sequencing (scRNA-seq) analysis of EMBI-enriched BM identified a small population (scRNA-seq cluster) of Μφs that specifically increased in number following administration of erythropoietin (Epo) concomitant with a similar increase in total BM EMBIs indicating this cluster, which expressed all of the defining EMBI Μφ genes such as Mertk, Dnase2a, and Hmox1, indeed consisted of EMBI Μφs. This population also exhibited strong and restricted expression of the heme-induced transcription factor Spic. To further explore the utility of Spic as a specific marker for EMBI Μφs, we examined EMBIs from Spicigfp/igfp mice which express GFP under the control of the Spic promoter and found that nearly 90% of the EMBI central Μφs were GFP+ indicating their expression of Spic. Moreover, evaluation of single BM cells by flow cytometry demonstrated that Spic-GFP+ Μφs are specifically enriched in the EMBI fraction and nearly 100% of the Spic-GFP expression in total BM cells belongs to the population of F4/80hi Μφs consistent with EMBI Μφs. Utilizing Spic-GFP as an endogenous marker for the EMBI Mφs, we were able to trypsinize whole BM from Spic-GFP mice stimulated with Epo or vehicle control (three biological replicates each) and the spleen of the Epo-treated mice (two biological replicates), and flow-sort the GFP+ cells for scRNAseq, avoiding Mφ fragmentation and reducing contamination from non-Mφ cells surrounded by Mφ membrane. This yielded more than 45,000 cells with the transcriptomic characteristics of classically-defined EMBI Mφs from eight 10x Chromium captures. Despite their many common EMBI Μφ genes, multiple distinct clusters could be identified in the murine BM at baseline and under stress erythropoiesis conditions, demonstrating the transcriptomic and surface marker heterogeneity that we and others have previously reported. Moreover, Epo stimulation led to differential gene expression within clusters, increases in certain Mφ populations, and the appearance of novel EMBI Μφ populations in both the BM and spleen during stress-erythropoiesis, indicating plasticity of existing macrophages and/or expansion of the EMBI Μφ pool via monocyte recruitment. We next examined the native structure of EMBIs in situ in Spicigfp/igfp mouse BM by confocal microscopy. Previous attempts to identify EMBIs in intact BM were typically confounded by the high density of cells in close proximity in space. Leveraging the expression of Spic-GFP to label the cell bodies, and powerful image analysis software (Imaris), we were able to visualize the thin lacy network of F4/80+ cytoplasmic extensions expanding as a scaffold throughout the marrow and cradling the Ter119+ erythroblasts and CD11b+ granulocyte precursor cells. The central Μφs and associated cells constituting an “island” span a much more extended area in situ than the small rosette of cells typically recognized as an erythro(myelo)blastic island in cytospins or by imaging flow cytometry, after typical mechanical fragmentation during BM extraction and preparation for studies. In order to examine these large extended EMBIs without disruption, we enzymatically dissociated spic-GFP+ BM with collagenase and DNAse and deposited the intact cell-clusters on clean glass coverslips; the Spic-GFP+ EMBI Mφs adhered, and each appeared to have at least 60-100 CD71+ and CD11b+ cells attached to their projections. These transcriptomic, imaging, and functional assay data implicate specialization of various EMBI Mφ populations to nurture erythropoiesis and granulopoiesis at different stages in order to have optimized production of blood cells at steady-state or stress conditions.

AbstractThe extraction of linguistic markers from social media posts, which are indicative of the onset and course of mental disorders, offers great potential for mental healthcare. In the present study, we extracted over one million posts from the popular social media platform Reddit to analyze speech coherence, which reflects formal thought disorder and is a characteristic feature of schizophrenia and associated psychotic disorders. Natural language processing (NLP) models were used to perform an automated quantification of speech coherence. We could demonstrate that users who are active on forums geared towards disorders with a higher degree of psychotic symptoms tend to show a lower level of coherence. The lowest coherence scores were found in users of forums on dissociative identity disorder, schizophrenia, and bipolar disorder. In contrast, a relatively high level of coherence was detected in users of forums related to obsessive–compulsive disorder, anxiety, and depression. Users of forums on posttraumatic stress disorder, autism, and attention-deficit hyperactivity disorder exhibited medium-level coherence. Our findings provide promising first evidence for the possible utility of NLP-based coherence analyses for the early detection and prevention of psychosis on the basis of posts gathered from publicly available social media data. This opens new avenues for large-scale prevention programs aimed at high-risk populations.

AimsCellular pathology (‘e-pathology’) record sets are a rich data resource with which to populate the electronic patient record (EPR). Accessible reports, even decades old, can be of great value in contemporary clinical decision making and as a resource for longitudinal clinical research. The aim of this short paper is to describe a solution in a major UK University Hospital which gives immediate visibility and clinical utility to 30 years of e-pathology recordsMethodsOver the past decade, we have created a timeline structured and iconographic data framework for the ‘whole-of-life’ visualisation of the entirety of an EPR. We have enhanced this interface with the sequential extraction of 373 342 e-pathology reports from legacy Ferranti (1990–1997) and Masterlab (1997–2004) files. They have been uploaded into our SQL file servers, following appropriate data quality and patient identity reconciliation checks.ResultsWe have restored a large repository of previously inaccessible e-pathology records to clinical use and to immediacy of access as a foundation element of our timeline structured EPR. This process has also allowed us to populate and validate an EPR-integral breast cancer data system of 20 000 cases with e-pathology records dating back to 1990.ConclusionsThe revitalisation of old e-pathology reports into a timeline structured EPR creates preserves and upcycles the investment in pathology reporting which is otherwise progressively lost to clinical use. E-pathology records provide reliable, life-long evidence of critical transition points in individual lives and disease progression for clinical and research use, when they can be instantly accessed.

IntroductionApplied practice experiences are essential components of the Masters of Public Health (MPH) curriculum. The objective of this study was to examine students’ perspectives on the skills and expertise they developed in an MPH course offering applied practice opportunities.MethodsOf 236 students who took the course from 2008 to 2018, email addresses were obtained for 212 and 104 completed the consent form. Following consent, reflection essays were de-identified and analyzed using a rapid qualitative analysis approach. The essays addressed students’ learning experiences and application of the competencies for MPH programs set by the Council for Education in Public Health (CEPH). Deductive and inductive analytical lenses were used to identify the key lessons learned by each cohort of students. Semi-structured guides and matrixes for essay analysis were created using assignment instructions and CEPH competencies.ResultsAlthough the reflection paper assignment varied across the years, commonalities were observed in the student reflections. Key themes included turning theory into practice, navigating the complex environment of public health practice, skill building, critical self-reflection, challenges encountered, and elements that facilitated project success. Students reported developing practical skills, such as planning for independent research (e.g., preparing for institutional review board (IRB) submission, consulting with faculty and other experts), identifying realistic approaches for data extraction during chart reviews and analyses of electronic medical records, and disseminating findings for diverse stakeholders and audiences. Students also reported strengthening cross-cutting skills such as communication, teamwork, and problem-solving that were useful for navigating power dynamics and balancing competing interests and expectations. Students explored their identity as public health professionals as they navigated the dynamics of public health practice.ConclusionThe applied practice experience served as a valuable tool for knowledge and skills acquisition. Moreover, it served as an opportunity for students to engage with the unique organizational cultures of their respective community partners and to deepen their understanding the complexities of conducting meaningful community-engaged research.ImplicationsThis study demonstrates the utility of analyzing students’ critical self-reflections as a tool for exploring learning experiences when training future public health professionals. The findings can help educators design future applied practice experiences.

Machilus thunbergii (Japanese bay tree) is native to warm temperate and subtropical regions in East Asia such as China, Japan, Korea, Taiwan, and Vietnam (Wu et al., 2006). This tree is used for landscape trees, windbreaks, and furniture because the wood is hard and dense (Hong et al., 2016). In May 2020, a leaf spot disease was observed on M. thunbergii in an arboretum on Wando Island, Korea. Among 25 trees surveyed in the arboretum, 7 trees showed 5 to 30% leaf spot disease. Symptoms consisted of gray and dry leaf spots up to approximately one to two centimeters in diameter, surrounded by a deep black margin. Leaf samples containing lesions were collected from the seven diseased trees. Pieces of leaf tissue (5mm × 5mm) were cut from the lesion margins and surface disinfected with 1% sodium hypochlorite (NaOCl) for 1 min and rinsed with sterile distilled water three times, patted dry on sterile paper towel and placed on Potato Dextrose Agar (PDA) in Petri dishes. From the cultures, ten fungal isolates were obtained and two representative isolates (CMML20-5 and CMML20-6) were stored at the Molecular Microbiology Laboratory, Chonnam National University, Gwangju, Korea. Colony morphology of the two isolates on PDA was observed after 7 days at 25°C in the dark. Conidiomata were induced after 7days in a 14h-10h light-dark condition using sufficiently grown mycelium in PDA, and both alpha and beta conidia were observed. Alpha conidia were 7.6 ± 0.9 × 2.8 ± 0.4 μm (n = 30), fusiform, aseptate, and hyaline. Beta conidia were 28.1 ± 3.6 × 2.7 ± 0.4 μm (n = 30), aseptate, hyaline, linear to hooked. Genomic DNA of the two isolates was extracted using the CTAB DNA extraction method (Cubero et al., 1999), followed by PCR using primer sets of the internal transcribed spacer (ITS1/ITS4) (White et al., 1990), elongation factor 1-α (EF1-728F/EF1-986R), calmodulin (CAL228F/CAL737R) (Carbone and Kohn, 1999), and TUB2 (Bt2a/Bt2b) (Glass and Donaldson 1995). PCR products were sequenced and analyzed to confirm species identity. The obtained sequences were deposited in GenBank (accession numbers OM049469, OM049470 for ITS, OM069429, OM069430 for EF1-α, OP130141, OP130142 for CAL, and OP130139, OP130140 for TUB2). BLASTn search analyses for ITS, EF1-α, CAL, and TUB2 sequences of two isolates selected resulted in near identical match (>97% for ITS, 100% for EF1-α, >99% for CAL, and >96% for TUB2) to sequences of Diaporthe eres strain AR4346 (=Phomopsis fukushii) (JQ807429 for ITS, JQ807355 for EF1-α, KJ435003 for CAL, and KJ420823 for TUB2). Phylogenetic analysis using maximum likelihood indicated that the two isolates grouped with reference strains (AR4346, AR4349, and AR4363) of D. eres with 76% bootstrap support. Based on morphological and phylogenetic analyses, the two isolates characterized in this study are members of the Diaporthe eres species complex as described by Udayanga et at. 2014. Pathogenicity tests were conducted using both detached leaf and whole plant assays. Mycelial PDA plugs (5-mm in diameter) or 10μl of 106 conidia suspensions were inoculated on detached leaves of M. thunbergii from 2-year-old trees and placed in 90 mm Petri-dishes containing wet filter papers or water agar medium. Mock inoculated controls used water in place of conidial suspensions. The plates were sealed with Parafilm and incubated at 25°C in the dark. Two year old M. thunbergii trees were inoculated with wet mycelia (1.5g) that was ground with a homogenizer and mixed with 50ml of sterile water and sprayed onto wounded leaves and stems with a needle. Mock inoculated controls were sprayed with water only. The inoculated seedlings were placed in plastic containers at 25 to 30°C to maintain high humidity. The pathogenicity tests were repeated three times with three replications. In detached leaves, symptoms of black spots were observed 6 days after mycelial plug inoculation and 20 days after conidia inoculation. In whole plants, typical symptoms were observed 9 days after inoculation. Symptoms were not observed on the control leaves and plants. Diaporthe eres was re-isolated from the inoculated leaf and whole plants and morphologically identified, fulfilling Koch’s postulates. Diaporthe eres has been reported to cause a leaf spot on Photinia × fraseri ‘Red Robin’ in China (Song et al. 2019). To our knowledge, this is the first report of leaf spot disease caused by Diaporthe eres on Japanese bay tree (Machilus thunbergii) in Korea. It is expected that use of this tree will expand given its utility, however infection with D. eres can cause serious diseases to the leaves and stems. Therefore, further studies on disease management are needed.

Cette thèse aborde plusieurs problèmes de sécurité et de confidentialité lors du traitement d'images médicales dans des lacs de données. Ainsi, on explore la fuite potentielle de données lors de l'exportation de modèles d'intelligence artificielle, puis on développe une approche d'anonymisation d'images médicales qui protège la confidentialité des données. Le Chapitre2 présente une nouvelle attaque d'exfiltration de données, appelée Data Exfiltration by Compression (DEC), qui s'appuie sur les techniques de compression d'images. Cette attaque est effectuée lors de l'exportation d'un réseau de neurones entraîné au sein d'un lac de données distant et elle est applicable indépendamment de la tâche de traitement d'images considérée. En explorant à la fois les méthodes de compression sans perte et avec perte, ce chapitre montre comment l'attaque DEC peut être utilisée efficacement pour voler des images médicales et les reconstruire avec une grande fidélité, grâce à l'utilisation de deux ensembles de données CT et IRM publics. Ce chapitre explore également les contre-mesures qu'un propriétaire de données peut mettre en œuvre pour empêcher l'attaque. Il étudie d'abord l'ajout de bruit gaussien pour atténuer cette attaque, et explore comment les attaquants peuvent créer des attaques résilientes à cet ajout. Enfin, une stratégie alternative d'exportation est proposée, qui implique un réglage fin du modèle et une vérification du code. Le Chapitre 3 présente une méthode d'anonymisation d'images médicales par approche générative, une nouvelle approche pour trouver un compromis entre la préservation de la confidentialité des patients et l'utilité des images générées pour résoudre les tâches de traitement d'images. Cette méthode sépare le processus d'anonymisation en deux étapes : tout d'abord, il extrait les caractéristiques liées à l'identité des patients et à l'utilité des images médicales à l'aide d'encodeurs spécialement entrainés ; ensuite, il optimise le code latent pour atteindre le compromis souhaité entre l'anonymisation et l'utilité de l'image. Nous utilisons des encodeurs d'identité, d'utilité et un encodeur automatique génératif basé sur un réseau antagoniste pour créer des images synthétiques réalistes à partir de l'espace latent. Lors de l'optimisation, nous incorporons ces encodeurs dans de nouvelles fonctions de perte pour produire des images qui suppriment les caractéristiques liées à l'identité tout en conservant leur utilité pour résoudre un problème de classification. L'efficacité de cette approche est démontrée par des expériences sur l'ensemble de données de radiographie thoracique MIMIC-CXR, où les images générées permettent avec succès la détection de pathologies pulmonaires. Le Chapitre 4 s'appuie sur les travaux du Chapitre 3 en utilisant des réseaux antagonistes génératifs (GAN) pour créer une solution d'anonymisation plus robuste et évolutive. Le cadre est structuré en deux étapes distinctes : tout d'abord, nous développons un encodeur simplifié et un nouvel algorithme d'entraînement pour plonger chaque image dans un espace latent. Dans la deuxième étape, nous minimisons les fonctions de perte proposées dans le Chapitre 3 pour optimiser la représentation latente de chaque image. Cette méthode garantit que les images générées suppriment efficacement certaines caractéristiques identifiables tout en conservant des informations diagnostiques cruciales. Des expériences qualitatives et quantitatives sur l'ensemble de données MIMIC-CXR démontrent que notre approche produit des images anonymisées de haute qualité qui conservent les détails diagnostiques essentiels, ce qui les rend bien adaptées à la formation de modèles d'apprentissage automatique dans la classification des pathologies pulmonaires. Le chapitre de conclusion résume les contributions scientifiques de ce travail et aborde les problèmes et défis restants pour produire des données médicales sensibles, sécurisées et préservant leur confidentialité
This thesis aims to address some specific safety and privacy issues when dealing with sensitive medical images within data lakes. This is done by first exploring potential data leakage when exporting machine learning models and then by developing an anonymization approach that protects data privacy.Chapter 2 presents a novel data exfiltration attack, termed Data Exfiltration by Compression (DEC), which leverages image compression techniques to exploit vulnerabilities in the model exporting process. This attack is performed when exporting a trained network from a remote data lake, and is applicable independently of the considered image processing task. By exploring both lossless and lossy compression methods, this chapter demonstrates how DEC can effectively be used to steal medical images and reconstruct them with high fidelity, using two public CT and MR datasets. This chapter also explores mitigation measures that a data owner can implement to prevent the attack. It first investigates the application of differential privacy measures, such as Gaussian noise addition, to mitigate this attack, and explores how attackers can create attacks resilient to differential privacy. Finally, an alternative model export strategy is proposed which involves model fine-tuning and code verification.Chapter 3 introduces the Generative Medical Image Anonymization framework, a novel approach to balance the trade-off between preserving patient privacy while maintaining the utility of the generated images to solve downstream tasks. The framework separates the anonymization process into two key stages: first, it extracts identity and utility-related features from medical images using specially trained encoders; then, it optimizes the latent code to achieve the desired trade-off between anonymity and utility. We employ identity and utility encoders to verify patient identities and detect pathologies, and use a generative adversarial network-based auto-encoder to create realistic synthetic images from the latent space. During optimization, we incorporate these encoders into novel loss functions to produce images that remove identity-related features while maintaining their utility to solve a classification problem. The effectiveness of this approach is demonstrated through extensive experiments on the MIMIC-CXR chest X-ray dataset, where the generated images successfully support lung pathology detection.Chapter 4 builds upon the work from Chapter 4 by utilizing generative adversarial networks (GANs) to create a more robust and scalable anonymization solution. The framework is structured into two distinct stages: first, we develop a streamlined encoder and a novel training scheme to map images into a latent space. In the second stage, we minimize the dual-loss functions proposed in Chapter 3 to optimize the latent representation of each image. This method ensures that the generated images effectively remove some identifiable features while retaining crucial diagnostic information. Extensive qualitative and quantitative experiments on the MIMIC-CXR dataset demonstrate that our approach produces high-quality anonymized images that maintain essential diagnostic details, making them well-suited for training machine learning models in lung pathology classification.The conclusion chapter summarizes the scientific contributions of this work, and addresses remaining issues and challenges for producing secured and privacy preserving sensitive medical data

The chapter focuses on the role of citizens and how this role is framed in official reports about waste. The hypothesis advanced is that the “objectivity” of the win-win game of waste recycling, accounted for by the official national reports, neutralizes the political overlapping of the neoliberal logic of waste commodification with the logics of public utility and urban welfare. A documentary analysis of official reports released by three of the most important actors performing Italian governance of waste management and recycling, is presented. The findings show how different logics adopted in the institutional accounting strategy contribute towards enacting a particular identity and agency of citizens in the process of waste valorisation.

Birch tree barks are regarded as waste in the pulp and papermaking industry and used as fuel. However, this material presents a source that contains many bio-based chemicals suitable for applications ranging from pharmaceuticals, plastics and composites, coatings, and antifeedants. Among the mixture of bio-derived chemicals in birch barks, triterpenoids, such as betulin, betulinic acid, and lupeol, can be present up to 30% weight of dry bark mass. They are highly valued for their anti-tumor, HIV, and inflammatory responses. In our presented work, triterpenoid mixtures were extracted through a Soxhlet extractor using the barks from locally sourced river birch trees (Betula nigra) with an average yield of 10.6% (dry bark mass). The extracted materials were characterized using the Nuclear Magnetic Resonance (NMR), Advanced Polymer Chromatography (APC), High-Performance Liquid Chromatography (HPLC), and hydroxyl number titration to assess the identity, average molecular weight, triterpenoid content, and the number of reactive sites, respectively. The extracts have been used to synthesize bio-based polymers with promising thermal and mechanical properties using minimal processing steps. Birch bark extract naturally contains many potential reactive sites and thus making it advantageous for synthesizing polymers without requiring multiple purification steps. We demonstrate the potentials for increasing the utility of birch bark, contributing to sustainability challenges in materials science and engineering.