Deep learning for cross-subject electroencephalography (EEG) decoding is hindered by high inter-subject variability, which introduces a severe domain shift between training and unseen test subjects. This survey presents a comprehensive review of deep learning methodologies specifically engineered to address this cross-subject generalization challenge. To ground this analysis, we formalize the cross-subject setting as a multi-source domain problem and delineate the rigorous, subject-independent evaluation protocols required for valid assessment. Central to this survey is a systematic taxonomy of the current literature into discrete methodological families, including feature alignment, adversarial learning, feature disentanglement, and contrastive learning. We conclude by examining three critical elements for advancing robust, real-world decoding: the theoretical limitations of current methodologies, the structural value of subject identity, and the emergence of EEG foundation models.

Spasticity, a type of hypertonia characterized by a velocity-dependent increase in muscle tone, is associated with damage to the brain and/or spinal cord in different neurological conditions. However, secondary non-neurological factors, such as soft tissue changes, can complicate the assessment and differentiation of the underlying causes. Accurate assessment is crucial for effective treatment planning, with clinicians relying on passive movement to grade the 'feel' of the spastic limb. This review aims to identify and evaluate the feasibility of spasticity models or simulations for clinical teaching. Models based on human spastic limbs were examined, with no restrictions on specific conditions or populations. A comprehensive search of four databases and gray literature was conducted to identify relevant studies. Criteria for inclusion focused on model development, data, and evaluation processes. Study selection and data extraction were carried out by independent reviewers, and data synthesis was performed by systematically mapping model properties, methods, and utility. The quality of the studies was assessed using an adapted framework for health technology assessments. The findings highlight opportunities for the development of simulation models to support training. However, significant limitations to the existing evidence base limit the feasibility of developing spasticity models based on existing literature.

Emerging evidence suggests that adipose tissue is not just a fat depot but a metabolically active organ that plays a central role in connecting obesity with its comorbidities. Understanding the complex interactions between adipocytes and neighboring cell types in obesity requires models that accurately replicate adipocyte behavior within their natural environment. Three-dimensional (3D) adipocyte cultures mimic the native tissue microenvironment by incorporating the spatial architecture as well as cell-cell and cell-extracellular matrix interactions presentin vivo, offering improved platforms for (patho)physiological adipose tissue modeling. 3D models of adipose tissue dysfunction enable the study of complex cellular crosstalk, such as adipocyte cancer cell interactions in breast, colorectal, bone, and pancreatic cancers; epicardial and pericardial adipocyte-myocardial cell dynamics in obesity-related cardiac dysfunction; and adipocyte-hepatocyte interactions in the development of non-alcoholic fatty liver disease, among other critical pathophysiological processes. In this review, we first discuss 3D models of adipose tissue and current strategies for mimicking the obesogenic microenvironment, including dietary stimulation of hyperlipidemia and hyperglycemia, as well as the incorporation of oxygen gradients, proinflammatory cytokines, and immune cells. Secondly, we examine 3D co-culture platforms that incorporate disease-associated/dysfunctional adipocytes with various cell types, such as cancer cells, cardiac cells, hepatocytes, immune cells, endothelial cells (EC), and fibroblasts, to model intercellular and interorgan crosstalk in obesity. Lastly, we provide insights into enhancing the physiological relevance of dysfunctional adipose tissue models and their co-culture systems while discussing future directions in tissue engineering aimed at improving clinical translation and reducing obesity related complications and mortality.

Peripheral Arterial Disease (PAD) is a common circulatory condition that leads to reduced blood flow to the limbs, often resulting in limb ischemia which can severely impact a patient's quality of life and increase the risk of amputation. Early diagnosis and timely intervention are critical in managing PAD-associated limb ischemia. This review provides a comprehensive overview of the latest diagnostic and therapeutic approaches for PAD and limb ischemia. We explored both non-invasive and invasive diagnostic techniques including ankle-brachial index (ABI), duplex ultrasonography, magnetic resonance angiography (MRA), computed tomography angiography (CTA) and emerging technologies like molecular imaging and near-infrared spectroscopy (NIRS). Therapeutic strategies discussed include pharmacological treatments such as antiplatelet agents and statins, endovascular interventions like angioplasty and stenting as well as advanced options such as gene and stem cell therapies. Emerging treatments including non-thermal plasma (NTP), and extracellular vesicle therapy are also highlighted for their regenerative potential. We have also addressed the challenges of current approaches, including diagnostic limitations, barriers to new therapies and cost considerations aimed at improving outcomes for PAD patients.&#xD.

Background.Transcatheter pulmonary valve replacement (TPVR) has emerged as a less invasive alternative to surgical pulmonary valve replacement for patients with right ventricular outflow tract dysfunction, such is especially important for those individuals whom had previous cardiac surgical procedures. Recently, three-dimensional (3D) printing and finite element (FE) computational simulation technologies have been employed to enhance preoperative planning processes; however, their effectiveness and clinical significance remain to be fully validated. This systematic review aims to describe the applications and potential impacts of 3D printing and FE simulation technologies for TPVR in clinical practice.Methods.A systematic search of PubMed, Science Direct, Web of Science, and Google Scholar was conducted to identify studies using patient-specific 3D-printed models and FE simulations for preoperative planning and device performance testing.Results.From 289 identified articles, 28 met the inclusion criteria for this review. The quality assessment of the articles showed that the article selection process was adequate. The eligible studies demonstrated that both 3D printing and FE-based simulations have been primarily used to select the appropriate pulmonary valve size as well as predict the optimal placement; i.e. to avoid potential complications such as paravalvular leakage or pulmonary regurgitation. These technologies are generally used in complex congenital and adult-congenital cases. Additionally, these studies provide valuable insights into the mechanical performances of the transcatheter valves using patient-specific anatomies.Conclusion.3D-printed models and FE simulations have both demonstrated utilities in TPVR planning; by accurately reproducing a given patient's anatomy and allowing evaluations of potential device-tissue interactions. These tools thus allow for personalized treatments and also contribute to device innovations and development. Yet, further research in this field is required due to the noted limitations of current studies, including small sample sizes, insufficient standardization, and/or challenges in replicating the biomechanics of cardiac tissue.

Image-guided tumor ablation (IGTA) has revolutionized modern oncological treatments by providing minimally invasive options that ensure precise tumor eradication with minimal patient discomfort. Traditional techniques such as ultrasound (US), computed tomography, and magnetic resonance imaging have been instrumental in the planning, execution, and evaluation of ablation therapies. However, these methods often face limitations, including poor contrast, susceptibility to artifacts, and variability in operator expertise, which can undermine the accuracy of tumor targeting and therapeutic outcomes. Incorporating deep learning (DL) into IGTA represents a significant advancement that addresses these challenges. This review explores the role and potential of DL in different phases of tumor ablation therapy: preoperative, intraoperative, and postoperative. In the preoperative stage, DL excels in advanced image segmentation, enhancement, and synthesis, facilitating precise surgical planning and optimized treatment strategies. During the intraoperative phase, DL supports image registration and fusion, and real-time surgical planning, enhancing navigation accuracy and ensuring precise ablation while safeguarding surrounding healthy tissues. In the postoperative phase, DL is pivotal in automating the monitoring of treatment responses and in the early detection of recurrences through detailed analyses of follow-up imaging. This review highlights the essential role of DL in modernizing IGTA, showcasing its significant implications for procedural safety, efficacy, and patient outcomes in oncology. As DL technologies continue to evolve, they are poised to redefine the standards of care in tumor ablation therapies, making treatments more accurate, personalized, and patient-friendly.

Optical coherence elastography (OCE) is a non-invasive imaging technique that measures the biomechanical properties of materials and tissues. This systematic review focuses on the applications of OCE in the anterior segment of the eye, including the cornea, iris, and crystalline lens, and its clinical relevance in diagnosing and managing ocular diseases. A systematic literature review was conducted using the PRISMA framework to identify studies published between 2014 and 2024. The review included studies that reported intrinsic biomechanical properties of anterior segment tissues measured using OCE. Databases searched included Scopus, Pub Med, and IEEE Xplore. Twenty-five studies met the inclusion criteria. The review found that OCE has been used to measure intrinsic biomechanical parameters such as Young's modulus and shear modulus in ocular tissues. OCE has been utilised to assess corneal stiffness in keratoconus, lens elasticity in presbyopia and cataract formation, and iris biomechanical changes under different lighting conditions. The studies demonstrated that OCE could detect subtle biomechanical changes associated with ocular diseases and measure treatment efficacy, such as collagen crosslinking for keratoconus management. The findings highlight the potential of OCE to enhance clinical diagnostics and patient care by providing detailed insights into the biomechanical properties of ocular tissues. However, variability in measurement techniques, the complexity of the method and reliance on animal models limit the current clinical translation of OCE. Standardised measurement protocols and further development andin vivovalidation are needed to overcome these barriers. OCE shows promise as a valuable non-invasive tool for high-resolution assessments of tissue biomechanics, which can subsequently support the diagnosis and management of ocular diseases. Future research should focus on standardising OCE methods and integrating them into clinical practice to fully realise their potential in improving patient outcomes.

Traumatic brain injuries (TBIs) pose a significant health concern among the elderly population, influenced by age-related physiological changes and the prevalence of neurodegenerative diseases. Understanding the biomechanical dimensions of TBIs in this demographic is vital for developing effective preventive strategies and optimizing clinical management. This comprehensive review explores the intricate biomechanics of TBIs in the elderly, integrating medical and aging studies, experimental biomechanics of head tissues, and numerical simulations. Research reveals that global brain atrophy in normal aging occurs at annual rates of -0.2% to -0.5%. In contrast, neurodegenerative diseases such as Alzheimer's, Parkinson's, and multiple sclerosis are associated with significantly higher rates of brain atrophy. These variations in atrophy rates underscore the importance of considering differing brain atrophy patterns when evaluating TBIs among the elderly. Experimental studies further demonstrate that age-related changes in the mechanical properties of critical head tissues increase vulnerability to head injuries. Numerical simulations provide insights into the biomechanical response of the aging brain to traumatic events, aiding in injury prediction and preventive strategy development tailored to the elderly. Biomechanical analysis is essential for understanding injury mechanisms and forms the basis for developing effective preventive strategies. By incorporating local atrophy and age-specific impact characteristics into biomechanical models, researchers can create targeted interventions to reduce the risk of head injuries in vulnerable populations. Future research should focus on refining these models and integrating clinical data to better predict outcomes and enhance preventive care. Advancements in this field promise to improve health outcomes and reduce injury risks for the aging population.

Retinal implantation of an electrode array is an emerging treatment for vision loss caused by outer-retinal degeneration. This article collects and analyses harms associated with the treatment reported in the peer-reviewed literature, thus enabling informed decision-making by patients, clinicians, researchers, engineers, and policymakers. We searched MEDLINE, Embase, and clinical trials registries for peer-reviewed journal articles reporting harms outcomes. We extracted data from articles including study design, definitions of 'serious adverse event', and timing of adverse events. We applied the McMaster tool to these articles to assess the risk of bias in harms assessment and reporting. Our searches returned 585 abstracts. We reviewed the full text of 59 articles describing 11 different devices. McMaster scores ranged from 3 to 12 (maximum 15; higher scores indicate less risk). We compiled a comprehensive list of all serious and non-serious adverse events associated with retinal implantation. Several harms were common across devices. Our meta-analysis showed that serious adverse events are log-uniformly distributed throughout follow-up. Improved reporting and further clinical studies are needed to develop a reliable safety profile of retinal implantation. Our findings will help guide the design, conduct, and reporting of future clinical trials of retinal implantation and other emerging treatments for vision loss. (PROSPERO registration: CRD42022308123.).

Mammography imaging remains the gold standard for breast cancer detection and diagnosis, but challenges in image quality can lead to misdiagnosis, increased radiation exposure, and higher healthcare costs. This comprehensive review evaluates traditional and machine learning-based techniques for improving mammography image quality, aiming to benefit clinicians and enhance diagnostic accuracy. Our literature search, spanning 2015 - 2024, identified 115 articles focusing on contrast enhancement and noise reduction methods, including histogram equalization, filtering, unsharp masking, fuzzy logic, transform-based techniques, and advanced machine learning approaches. Machine learning, particularly architectures integrating denoising autoencoders with convolutional neural networks, emerged as highly effective in enhancing image quality without compromising detail. The discussion highlights the success of these techniques in improving mammography images' visual quality. However, challenges such as high noise ratios, inconsistent evaluation metrics, and limited open-source datasets persist. Addressing these issues offers opportunities for future research to further advance mammography image enhancement methodologies.