Medical image denoising is considered among the most challenging vision tasks. Despite the real-world implications, existing denoising methods have notable drawbacks as they often generate visual artifacts when applied to heterogeneous medical images. This study addresses the limitation of the contemporary denoising methods with an artificial intelligence (AI)-driven two-stage learning strategy. The proposed method learns to estimate the residual noise from the noisy images. Later, it incorporates a novel noise attention mechanism to correlate estimated residual noise with noisy inputs to perform denoising in a course-to-refine manner. This study also proposes to leverage a multi-modal learning strategy to generalize the denoising among medical image modalities and multiple noise patterns for widespread applications. The practicability of the proposed method has been evaluated with dense experiments. The experimental results demonstrated that the proposed method achieved state-of-the-art performance by significantly outperforming the existing medical image denoising methods in quantitative and qualitative comparisons. Overall, it illustrates a performance gain of 7.64 in Peak Signal-to-Noise Ratio (PSNR), 0.1021 in Structural Similarity Index (SSIM), 0.80 in DeltaE ($ΔE$), 0.1855 in Visual Information Fidelity Pixel-wise (VIFP), and 18.54 in Mean Squared Error (MSE) metrics.
Felix Buendía, Joaquín Gayoso-Cabada, José-Luis Sierra
In this paper, we describe an approach to transforming the huge amount of medical knowledge available in existing online medical collections into standardized learning packages ready to be integrated into the most popular e-learning platforms. The core of our approach is a tool called Clavy, which makes it possible to retrieve pieces of content in medical collections, to transform this content into meaningful learning units, and to export it in the form of standardized learning packages. In addition to describing the approach, we demonstrate its feasibility by applying it to the generation of IMS content packages from MedPix, a popular online database of medical cases in the domain of radiology.
Nikita Malik, Pratinav Seth, Neeraj Kumar Singh
et al.
Deep learning has driven significant advances in medical image analysis, yet its adoption in clinical practice remains constrained by the large size and lack of transparency in modern models. Advances in interpretability techniques such as DL-Backtrace, Layer-wise Relevance Propagation, and Integrated Gradients make it possible to assess the contribution of individual components within neural networks trained on medical imaging tasks. In this work, we introduce an interpretability-guided pruning framework that reduces model complexity while preserving both predictive performance and transparency. By selectively retaining only the most relevant parts of each layer, our method enables targeted compression that maintains clinically meaningful representations. Experiments across multiple medical image classification benchmarks demonstrate that this approach achieves high compression rates with minimal loss in accuracy, paving the way for lightweight, interpretable models suited for real-world deployment in healthcare settings.
Image denoising algorithms have been extensively investigated for medical imaging. To perform image denoising, penalized least-squares (PLS) problems can be designed and solved, in which the penalty term encodes prior knowledge of the object being imaged. Sparsity-promoting penalties, such as total variation (TV), have been a popular choice for regularizing image denoising problems. However, such hand-crafted penalties may not be able to preserve task-relevant information in measured image data and can lead to oversmoothed image appearances and patchy artifacts that degrade signal detectability. Supervised learning methods that employ convolutional neural networks (CNNs) have emerged as a popular approach to denoising medical images. However, studies have shown that CNNs trained with loss functions based on traditional image quality measures can lead to a loss of task-relevant information in images. Some previous works have investigated task-based loss functions that employ model observers for training the CNN denoising models. However, such training processes typically require a large number of noisy and ground-truth (noise-free or low-noise) image data pairs. In this work, we propose a task-based regularization strategy for use with PLS in medical image denoising. The proposed task-based regularization is associated with the likelihood of linear test statistics of noisy images for Gaussian noise models. The proposed method does not require ground-truth image data and solves an individual optimization problem for denoising each image. Computer-simulation studies are conducted that consider a multivariate-normally distributed (MVN) lumpy background and a binary texture background. It is demonstrated that the proposed regularization strategy can effectively improve signal detectability in denoised images.
Semi-supervised 3D medical image segmentation aims to achieve accurate segmentation using few labelled data and numerous unlabelled data. The main challenge in the design of semi-supervised learning methods consists in the effective use of the unlabelled data for training. A promising solution consists of ensuring consistent predictions across different views of the data, where the efficacy of this strategy depends on the accuracy of the pseudo-labels generated by the model for this consistency learning strategy. In this paper, we introduce a new methodology to produce high-quality pseudo-labels for a consistency learning strategy to address semi-supervised 3D medical image segmentation. The methodology has three important contributions. The first contribution is the Cooperative Rectification Learning Network (CRLN) that learns multiple prototypes per class to be used as external knowledge priors to adaptively rectify pseudo-labels at the voxel level. The second contribution consists of the Dynamic Interaction Module (DIM) to facilitate pairwise and cross-class interactions between prototypes and multi-resolution image features, enabling the production of accurate voxel-level clues for pseudo-label rectification. The third contribution is the Cooperative Positive Supervision (CPS), which optimises uncertain representations to align with unassertive representations of their class distributions, improving the model's accuracy in classifying uncertain regions. Extensive experiments on three public 3D medical segmentation datasets demonstrate the effectiveness and superiority of our semi-supervised learning method.
Objetivo: Obtener el perfil genético de diferentes tejidos de dos individuos sometidos a embalsamamiento con un tiempo aproximado de ocho meses.
Introducción: Las técnicas de embalsamamiento buscan la preservación temporal de los tejidos después del deceso mediante la aplicación de sustancias químicas, a través de la red de vasos sanguíneos; llenando la cavidad torácica y abdominal o directamente embebiendo partes del cuerpo. Estas sustancias ocasionan daños a la integridad del material genético, por ejemplo, el formaldehido, ampliamente utilizado en la tanatopraxia, puede llegar a causar entrecruzamiento de los ácidos nucleicos y proteínas; otro ejemplo son los colorantes, identificados como inhibidores de la reacción en cadena de la polimerasa (por sus siglas en inglés PCR). Obtener material genético de muestras problema para la identificación humana en términos de impartición de justicia, requiere de metodologías que optimicen el análisis en todas sus etapas; la cantidad y tipo de muestra, insumos y temporalidades.
Metodología: A partir de dos individuos sometidos a embalsamamiento desde hace aproximadamente ocho meses, se seleccionaron 33 muestras de: tejido muscular visiblemente putrefacto, cartílago, hueso compacto, hueso esponjoso de la médula y la médula ósea amarilla (huesos largos), la cual fue embebida en tarjetas QIAcard FTA. Las tarjetas se lavaron con QIAcard FTA Wash Buffer, mientras que las muestras de tejido, cartílago y hueso se trataron con los reactivos de QIAquick Gel Extraction Kit y PrepFiler® BTA Forensic DNA Extraction Kit (con el equipo AutoMate Express™ Forensic DNA Extraction System). La amplificación del DNA genómico fue mediante PCR, utilizando PowerPlex® Fusion 6C System, se identificaron los alelos por electroforesis capilar en el equipo Genetic Analyzer 3500 y finalmente los datos del corrimiento se analizaron mediante el software GeneMapper ID-X.
Resultados: Las muestras de médula ósea amarilla embebida en tarjeta QIAcard FTA resultaron con mayor eficiencia en la recuperación de ADN, seguidas de las muestras óseas y el tejido muscular, extraídas con perlas magnéticas y membranas de sílice, siendo el cartílago el que presenta mayor alteración del ADN debido a la exposición directa a las sustancias químicas utilizadas en el proceso de embalsamamiento. En total se obtuvieron 16 perfiles completos, 10 perfiles parciales y 7 nulos, posibilitando la identificación de ambos individuos en un periodo corto de tiempo.
Discusión y conclusión: Con las metodologías utilizadas se obtuvieron perfiles genéticos de diferentes tejidos sometidos a embalsamamiento, suficiente para la identificación de los individuos. Sin embargo, la médula ósea amarilla constituida principalmente por grasa y células madre que se embebió en tarjeta QIAcard FTA permitió acelerar el proceso de extracción y purificación de ADN comparada con las muestras de tejido muscular, cartílago y óseo (compacto y esponjoso) que conllevan un proceso de muestreo y extracción con un tiempo considerablemente mayor, evidenciando que, un plan adecuado de muestreo es clave para la obtención de resultados, optimizando la estandarización de los protocolos y kits forenses del laboratorio.
O objetivo deste artigo foi descrever as características pelas quais o desenvolvimento da inteligência artificial em aplicações clínicas não pode ser isolado do processo geral seguido por outros mecanismos de extração de renda já praticados nas plataformas de uso de tecnologia digital em outros domínios. A metodologia utilizada foi a de revisão sintética da literatura recente, acadêmica e jornalística, analisando os mecanismos de relação entre a concentração de renda e a degradação algorítmica em diversos campos, incluindo o da saúde. Os resultados levantados mostram que esses mecanismos levam, de maneira inevitável, ao ajuste ou sintonia dos algoritmos e à seleção dos resultados a fim de produzir os melhores retornos para o capital. A lógica por trás das corporações tecnológicas conduz, inexoravelmente, a um processo progressivo, no qual os melhores interesses dos pacientes devem ser pospostos sempre que existam boas razões comerciais. Esse processo independe das boas intenções e do altruísmo das startups da saúde. Concluiu-se que, sem participação ativa do Estado, o ecossistema de geração de soluções digitais para a saúde vai ter uma degradação inevitável para privilegiar a extração de renda, prejudicando os interesses dos pacientes.
Law, Law in general. Comparative and uniform law. Jurisprudence
Nora Ellen Groce, Samia Hurst, Laura Catalina Izquierdo Martínez
et al.
Objectives To identified the core components of professional ethics for medical sign language interpreters and develop a framework based on empirical data from Colombian sign language (CSL) interpreters.Design Purposive and snowball sampling methods were used in this qualitative study, which involved conducting semistructured interviews to CSL interpreters. Inductive data analysis was performed using the constant comparative method, where data collection and analysis occurred simultaneously. Transcriptional coding was performed line by line, and the data results were organised into themes and subthemes.Setting The research was conducted in Colombia.Participants A total of 17 CSL interpreters were included.Results We identified key themes (confidentiality, privacy, professionalism, business practices and professional development). Our data analyses show the need for codes of conduct and establishment of professional codes of ethics for sign language interpreters working in a health context. The proposed framework addresses the challenges within the professional ethics of sign language interpreters in healthcare.Conclusions These findings offer unique insights into the ethical experiences of CSL interpreters. This framework can be a valuable reference for interpreters facing ethical dilemmas. The clarity of ethical considerations is crucial for overcoming barriers to healthcare for the D/deaf population. The identified ethical issues underscore the necessity of education, training and the establishment of codes of ethics and legislation for sign language interpreters. These findings can serve as a foundational reference for crafting ethical guidelines for sign language interpreters in other low- and middle-income countries.
Johanna Korfitsen, Eva Samuelsson, David Forsström
et al.
To strengthen the right to support for people with gambling problems in Sweden, legislative changes were enacted in 2018. This study aims to critically examine how problems and solutions are represented in 69 appeals concerning gambling treatment within the general administrative court (2014–2022) and to assess how these representations have evolved following the legal amendments. The study employs Bacchi’s WPR approach to scrutinize court judgments. The results reveal that gambling problems are unequivocally recognized as severe issues requiring intervention, with both explicit and implicit notions of the problem rooted in the concept of loss of control. Prior to the legal amendments, rulings primarily focused on identifying the responsible actor for providing care, often framed within a medical discourse. Post-amendment, the focus shifted to how treatment needs should be met, emphasizing an evidence-based discourse. These varying representations produce discursive, subjectifying, and material consequences, significantly affecting access to different welfare interventions. The new legislation has solidified the responsibility of social services to provide treatment for gambling problems. However, as the study demonstrates, responsibilization of gamblers occurs not only in policy and treatment frameworks, but also within the court system.
Medical image representations can be learned through medical vision-language contrastive learning (mVLCL) where medical imaging reports are used as weak supervision through image-text alignment. These learned image representations can be transferred to and benefit various downstream medical vision tasks such as disease classification and segmentation. Recent mVLCL methods attempt to align image sub-regions and the report keywords as local-matchings. However, these methods aggregate all local-matchings via simple pooling operations while ignoring the inherent relations between them. These methods therefore fail to reason between local-matchings that are semantically related, e.g., local-matchings that correspond to the disease word and the location word (semantic-relations), and also fail to differentiate such clinically important local-matchings from others that correspond to less meaningful words, e.g., conjunction words (importance-relations). Hence, we propose a mVLCL method that models the inter-matching relations between local-matchings via a relation-enhanced contrastive learning framework (RECLF). In RECLF, we introduce a semantic-relation reasoning module (SRM) and an importance-relation reasoning module (IRM) to enable more fine-grained report supervision for image representation learning. We evaluated our method using six public benchmark datasets on four downstream tasks, including segmentation, zero-shot classification, linear classification, and cross-modal retrieval. Our results demonstrated the superiority of our RECLF over the state-of-the-art mVLCL methods with consistent improvements across single-modal and cross-modal tasks. These results suggest that our RECLF, by modelling the inter-matching relations, can learn improved medical image representations with better generalization capabilities.
Melis Selamoglu, Bircan Erbas, Hester Wilson
et al.
Abstract Background A significant policy change impacting the availability of nicotine for use in electronic cigarettes (e-cigarettes) in Australia took effect from October 1, 2021. This change meant that nicotine containing liquids for use with e-cigarettes would only be available by prescription from a medical practitioner as part of a smoking cessation plan. This study aimed to explore general practitioners (GPs) perceptions about the role of e-cigarettes, and understand factors informing their intentions to prescribe e-cigarettes as part of a smoking cessation plan. Methods In-depth semi-structured interviews were conducted with thirteen GPs. Purposeful sampling was used to recruit participants. Interviews were audio recorded and transcribed verbatim. Thematic analysis was used to classify, describe and report themes in the data. QSR NVivo was used to aid coding, thematic analysis and retrieval of quotes. Results Participants had diverse views on recommending and prescribing e-cigarettes as smoking cessation aids to patients. Some participants were willing to prescribe e-cigarettes to patients if other methods of smoking cessation had not worked but there were concerns, and uncertainty, about the safety and efficacy of e-cigarettes for smoking cessation. There was poor understanding of the current policy and legislation about e-cigarettes in Australia. Mostly the participants in this sample did not feel confident or comfortable to prescribe, or have discussions about e-cigarettes with patients. Conclusion The participants of this study held diverse attitudes on recommending and prescribing e-cigarettes for smoking cessation. Clarity in guidelines and consumer product information are required to enable GPs to provide consistent and accurate advice to patients that wish to use e-cigarettes as a smoking cessation aid.
Christos Matsoukas, Johan Fredin Haslum, Moein Sorkhei
et al.
Convolutional Neural Networks (CNNs) have reigned for a decade as the de facto approach to automated medical image diagnosis, pushing the state-of-the-art in classification, detection and segmentation tasks. Over the last years, vision transformers (ViTs) have appeared as a competitive alternative to CNNs, yielding impressive levels of performance in the natural image domain, while possessing several interesting properties that could prove beneficial for medical imaging tasks. In this work, we explore the benefits and drawbacks of transformer-based models for medical image classification. We conduct a series of experiments on several standard 2D medical image benchmark datasets and tasks. Our findings show that, while CNNs perform better if trained from scratch, off-the-shelf vision transformers can perform on par with CNNs when pretrained on ImageNet, both in a supervised and self-supervised setting, rendering them as a viable alternative to CNNs.
Medical Vision-Language Pre-training (VLP) learns representations jointly from medical images and paired radiology reports. It typically requires large-scale paired image-text datasets to achieve effective pre-training for both the image encoder and text encoder. The advent of text-guided generative models raises a compelling question: Can VLP be implemented solely with synthetic images generated from genuine radiology reports, thereby mitigating the need for extensively pairing and curating image-text datasets? In this work, we scrutinize this very question by examining the feasibility and effectiveness of employing synthetic images for medical VLP. We replace real medical images with their synthetic equivalents, generated from authentic medical reports. Utilizing three state-of-the-art VLP algorithms, we exclusively train on these synthetic samples. Our empirical evaluation across three subsequent tasks, namely image classification, semantic segmentation and object detection, reveals that the performance achieved through synthetic data is on par with or even exceeds that obtained with real images. As a pioneering contribution to this domain, we introduce a large-scale synthetic medical image dataset, paired with anonymized real radiology reports. This alleviates the need of sharing medical images, which are not easy to curate and share in practice. The code and the dataset can be found in \href{https://github.com/cheliu-computation/MedSyn-RepLearn/tree/main}{https://github.com/cheliu-computation/MedSyn-RepLearn/tree/main}.
The idea of using computers to read medical scans was introduced as early as 1966. However, limits to machine learning technology meant progress was slow initially. The Alexnet breakthrough in 2012 sparked new interest in the topic, which resulted in the release of 100s of medical AI solutions on the market. In spite of success for some diseases and modalities, many challenges remain. Research typically focuses on the development of specific applications or techniques, clinical evaluation, or meta analysis of clinical studies or techniques through surveys or challenges. However, limited attention has been given to the development process of improving real world performance. In this tutorial, we address the latter and discuss some techniques to conduct the development process in order to make this as efficient as possible.
Trusting the predictions of deep learning models in safety critical settings such as the medical domain is still not a viable option. Distentangled uncertainty quantification in the field of medical imaging has received little attention. In this paper, we study disentangled uncertainties in image to image translation tasks in the medical domain. We compare multiple uncertainty quantification methods, namely Ensembles, Flipout, Dropout, and DropConnect, while using CycleGAN to convert T1-weighted brain MRI scans to T2-weighted brain MRI scans. We further evaluate uncertainty behavior in the presence of out of distribution data (Brain CT and RGB Face Images), showing that epistemic uncertainty can be used to detect out of distribution inputs, which should increase reliability of model outputs.
Contrastive learning has shown great promise over annotation scarcity problems in the context of medical image segmentation. Existing approaches typically assume a balanced class distribution for both labeled and unlabeled medical images. However, medical image data in reality is commonly imbalanced (i.e., multi-class label imbalance), which naturally yields blurry contours and usually incorrectly labels rare objects. Moreover, it remains unclear whether all negative samples are equally negative. In this work, we present ACTION, an Anatomical-aware ConTrastive dIstillatiON framework, for semi-supervised medical image segmentation. Specifically, we first develop an iterative contrastive distillation algorithm by softly labeling the negatives rather than binary supervision between positive and negative pairs. We also capture more semantically similar features from the randomly chosen negative set compared to the positives to enforce the diversity of the sampled data. Second, we raise a more important question: Can we really handle imbalanced samples to yield better performance? Hence, the key innovation in ACTION is to learn global semantic relationship across the entire dataset and local anatomical features among the neighbouring pixels with minimal additional memory footprint. During the training, we introduce anatomical contrast by actively sampling a sparse set of hard negative pixels, which can generate smoother segmentation boundaries and more accurate predictions. Extensive experiments across two benchmark datasets and different unlabeled settings show that ACTION significantly outperforms the current state-of-the-art semi-supervised methods.
<h4>Background</h4>Studies have indicated that the advance care planning knowledge and attitudes of elderly individuals strongly affect their implementation of advance care planning. A measurement with a theoretical base for evaluating elderly individuals' knowledge, attitudes, and behavioral intentions regarding advance care planning is lacking.<h4>Objectives</h4>To develop a questionnaire and understand elderly individuals' knowledge, attitudes, and behavioral intentions regarding implementing advance care planning.<h4>Methods</h4>A cross-sectional questionnaire survey was conducted. The content validity index, and statistical methods, including discrimination, factor, and reliability analysis, were adopted for psychometric testing. Descriptive statistics mainly presented data analysis.<h4>Results</h4>401 elderly individuals were recruited from a medical center and one senior activity center. The content validity index was approximately 0.71-0.92 for the developed questionnaires, the Kuder-Richardson formula 20 was 0.84 for advance care planning knowledge, and the Cronbach's alpha was 0.86, 0.94, 0.76, and 0.92 for attitudes, behavioral intentions, influencing factors, and subjective norms, respectively. The average score for advance care planning knowledge for elderly individuals was 4.42, with a correct answer rate of 49.1%. They lacked knowledge of advance care planning-related legislation. The mean score for attitudes and behavioral intentions was 14.32 and 3.48, respectively. Elderly individuals agreed that advance care planning has benefits but were worried about the emotional distress caused by advance care planning discussions. Elderly individuals with positive behavioral intentions tend to implement advance care planning. Spouses, children, doctors, and nurses are significant reference people for elderly individuals.<h4>Conclusions</h4>The developed questionnaire exhibits good validity and reliability for understanding elderly individuals' knowledge, attitudes, and behavioral intentions concerning advance care planning. Advance care planning materials or decision aids suitable for elderly individuals must be developed to increase their understanding of advance care planning. Additionally, the role of nurses is indispensable in promoting advance care planning among elderly individuals.
The purpose of the study is to analyze the current labor, criminal and criminal procedure legislation, which was adopted in the martial law and events occurring in Ukraine.
 Materials and Methods. The analysis of normative-legal acts adopted from
 February 24, 2022, directly related to the rules and responsibilities of citizens of Ukraine, the work of public institutions in martial law. The provisions of labor, criminal and criminal procedure legislation of Ukraine are substantiated, which are extremely necessary during the investigation of criminal proceedings that are important for citizens of Ukraine, institutions of any form of ownership, working and continuing to implement state policy in hostilities.
 Results and conclusions. The provisions of labor, criminal and criminal procedure legislation of Ukraine are substantiated, which are extremely necessary during the investigation of criminal proceedings that are important for citizens of Ukraine, institutions of any form of ownership, working and continuing to implement state policy in hostilities.
 The procedure for conducting investigative actions by authorized persons, violation of current legislation and realization of the right to protection have been determined on the basis of study and analysis of legislative norms.
The medical device products at the European Union market must be safe and effective. To ensure this, medical device manufacturers must comply to the new regulatory requirements brought by the Medical Device Regulation (MDR) and the In Vitro Diagnostic Medical Device Regulation (IVDR). In general, the new regulations increase regulatory requirements and oversight, especially for medical software, and this is also true for requirements related to cybersecurity, which are now explicitly addressed in the legislation. The significant legislation changes currently underway, combined with increased cybersecurity requirements, create unique challenges for manufacturers to comply with the regulatory framework. In this paper, we review the new cybersecurity requirements in the light of currently available guidance documents, and pinpoint four core concepts around which cybersecurity compliance can be built. We argue that these core concepts form a foundations for cybersecurity compliance in the European Union regulatory framework.