Synthetic data generation plays a crucial role in medical research by mitigating privacy concerns and enabling large-scale patient data analysis. This study presents a Graph Convolutional Neural Network combined with a Beta-Variational Autoencoder (GCN-β-VAE) framework for generating synthetic Abdominal Aortic Aneurysms (AAA). Using a small real-world dataset, our approach extracts key anatomical features and captures complex statistical relationships within a compact disentangled latent space. To address data limitations, low-impact data augmentation based on Procrustes analysis was employed, preserving anatomical integrity. The generation strategies, both deterministic and stochastic, manage to enhance data diversity while ensuring realism. Compared to PCA-based approaches, our model performs more robustly on unseen data by capturing complex, nonlinear anatomical variations. This enables more comprehensive clinical and statistical analyses than the original dataset alone. The resulting synthetic AAA dataset preserves patient privacy while providing a scalable foundation for medical research, device testing, and computational modeling.

Audit and re-audit regarding the current practice of ABCDE approach in the emergency department at selected university hospitals in Syria.

Objectives: This study aimed to assess patient satisfaction with oral healthcare based on accessibility, care environment and care quality. Methods: To meet the above objective, a descriptive cross-sectional study was conducted among patients visiting the Biyem-Assi (BA) and Cité Verte District Hospitals in Yaoundé, Cameroon, from 2021 to 2022. Satisfaction levels were assessed using an administered questionnaire derived from the Dental Satisfaction questionnaire (DSQ) and the Service Quality Questionnaire (SERVQUAL). Data were analysed using the Statistical Package for Social Sciences (SPSS) version 25. The mean was used for quantitative data, and the Likert scale was used to assess qualitative data. Results with p-values less than 5 were considered statistically significant. Results: Two hundred participants were included in this study. The age range from 29 to 39 years was the most represented (40%), and most participants were from the grassfields (41.5%). The literacy rate was high, 73.1% of participants had a tertiary level of education. Satisfaction with access to oral health in the study hospitals was 67.5%. Fifty-seven per cent of participants reported satisfaction with the hospital environment. Almost all participants (98%) were satisfied with the quality of care received. A global satisfaction of 71% was recorded. The reasons for patient dissatisfaction included high treatment costs (41%), lack of patient intimacy (60%), and long waiting hours (60-300 minutes). Conclusion: Patient satisfaction with oral healthcare was high. Major complaints included the non-respect of patient privacy, long waiting hours and high cost of treatment.

Purpose Diabetic retinopathy is an ophthalmic disease that impairs the retinal blood vessels. Diabetic retinopathy can lead to blindness when it is not examined in earlier phases. Adversely, the accurate diabetic retinopathy recognition phase is prominently complicated and needs experienced human analysis of fundus images. Blockchain technology helps share data by allowing users to select what information to share and control who can access it, which is important for managing electronic health records in healthcare sector. Nevertheless, the privacy of user data is compromised due to the training data, which is revealed to unauthorized users. Methods In this work, a superior module for diabetic retinopathy classification based on Blockchain using principal convolutional analysis neural network is designed. Here, the simulation of Blockchain is carried out. Here, the input image is pre-processed using the Gaussian filter. LadderNet is deployed for lesion segmentation, and the segmentation of blood vessel is done using the Sine-Net model. Moreover, feature extraction is performed with the input image, lesion-segmented image, and blood vessel-segmented image. Finally, diabetic retinopathy classification is executed utilizing the proposed principal convolutional analysis neural network, which is classified into normal, mild non-proliferative diabetic retinopathy, moderate non-proliferative diabetic retinopathy, severe non-proliferative diabetic retinopathy, and proliferative. Results The Blockchain enabled principal convolutional analysis neural network obtains superior values of 90.9%, 91.9%, 92.5%, 89.4%, 88.4%, and 75.5% in terms of metrics like accuracy, true positive rate, true negative rate, positive predictive value, negative predictive value, and Mathews correlation coefficient. Conclusion The integration of principal convolutional analysis neural network with Blockchain enhances data integrity and patient privacy, making it a promising solution for early diagnosis and treatment. Also, this approach ensures accurate and efficient detection of diabetic retinopathy.

The U.S. healthcare system has been experiencing an increasingly cybersecurity crisis, with more than 276 million individuals having had their data stolen in 2024 and hacking-related breaches accounting for almost 80% of reported incidents. Traditional security measures have failed to combat advanced cyber threats that seek to steal sensitive PII, health insurance information, and medical infrastructure. This study analyzes the creation, deployment and operational outcomes of AI-driven cyber-risk prediction models for U.S. healthcare institutions. A systematic literature review (SLR) design was used to analyze peer-reviewed academic journals and cybersecurity reports between 2012 to 2025 in databases such as IEEE Xplore, PubMed and ACM Digital Library. The study compares machine learning approaches for supervising the learning process, deep learning models and natural language processing tasks in healthcare cybersecurity. The findings indicated that AI models show superior performance in threat and risk prediction, particularly with gradient boosting algorithms, which yield the best accuracy for vulnerability identification. However, there continue to be barriers to implementation such as resource limitations, existing infrastructure needs, workforce skill gaps and regulatory uncertainty. Budget allocation is found to be the most important determinant of AI adoption success. AI technology has the potential to transform health care cybersecurity, but true investment value will only be seen when the necessary infrastructures are in place through strategic investments, training of staff and people with IT expertise, policy harmonization (both regulatory and policy), inclusive collaborative frameworks that support collective defense and are HIPAA compliant, while taking into account patient privacy issues. Keywords: Artificial Intelligence, Cyber Risk, Prediction Models, US Healthcare Institutions

Background and Objectives: Telehealth can improve pediatric access to care; however, substantial telehealth disparities continue to persist. Publications describing the implementation of innovative strategies to improve telehealth equity across pediatric health care organizations are limited. We aimed to describe the degree of implementation of telehealth equity interventions, identify facilitators and barriers to their implementation, and develop a quality improvement framework for future implementation efforts. Methods: We conducted a national survey within the American Academy of Pediatrics' Supporting Pediatric Research on Outcomes and Utilization of Telehealth network from October to December 2023 to obtain an environmental scan of telehealth equity interventions adopted by pediatric health organizations since the onset of the COVID-19 pandemic. We used descriptive statistics to quantify intervention implementation and employed qualitative template analysis of open-ended responses to design a key driver diagram blueprint to advance pediatric telehealth equity. Results: Of 134 organizations, members from 13 organizations completed the survey, mostly representing large, urban, academic children's hospitals. The most commonly implemented interventions included providing devices or internet access within health care facilities, offering on-demand help desk support, and integrating interpreter services within telehealth visit platforms. In contrast, less commonly implemented interventions included expanding broadband access in community settings, soliciting family visit preferences for telehealth versus in-person visits, and ensuring patient privacy during telehealth encounters. Conclusions: This study describes the degree of implementation of telehealth equity interventions among well-resourced pediatric health organizations. Our quality improvement framework provides a foundation for future multicenter, collaborative initiatives aimed at reducing telehealth disparities in pediatric populations.

Healthcare 5.0 shows an evolution in intelligent, integrated, and patient-centric health systems, which is based on the interconnected use of technologies such as Internet of Medical Things (IoMT), Artificial Intelligence (AI) and edge computing. But, with increasing dependence on these technologies, it also causes several concers, such as data privacy, security, and interoperability. The sharing of data at one centre for training can increase the risk of data breach, raising worries about data safety. Federated learning, on the other hand, allows training of machine learning models without sharing data in a decentralised manner. The patient privacy and data remain protected as sensitive data can remain on the device. Simultaneously, Blockchain improves this framework by providing a secure and transparent record of data transactions, providing accurate data tracking which further reduces the risk of data tampering. In this paper, we will evaluate the integration of Federated learning with blockchain technology in the framework of Internet of Medical Things (IoMT). We present a system design which guarantees secure data exchange, reliable model aggregation and traceability. Additionally, We will also discuss about challenges and ethical consideration in its implementation.

Objective.With the advancement of deep learning technologies, more and more researchers have begun developing end-to-end automatic sleep stage classification frameworks. However, these frameworks typically require access to large electroencephalogram (EEG) datasets for training, which imposes a significant computational burden. Furthermore, EEG data contains patient privacy information, and using such data for training raises concerns about privacy infringement. To address these issues, we propose a hybrid data distillation method. We aim to enable single-channel EEG sleep stage classification with less training cost and privacy risk by distilling large real datasets into a tiny, privacy-preserving synthetic set for training from scratch.Approach.We first apply the gradient matching method to optimize the randomly initialized synthetic dataset. The gradient changes in the early stages of model training can quickly reduce the performance gap between the synthetic dataset and the source dataset. Subsequently, to avoid oscillations near the optimal solution during gradient matching, we switch to distribution matching to further optimize the synthetic dataset. This method aligns the data distribution at a global level, enhancing overall consistency. In addition, we adopt a novel mini-batch iteration method to assist the synthetic dataset in learning temporal dependencies.Main results.We validated our framework on three public datasets and achieved robust results.Significance.This study proposes an efficient and robust hybrid data distillation algorithm, providing a feasible approach for implementing sleep stage staging based on privacy protection.

In the contemporary era of healthcare informatics, preserving the privacy of EHD while ensuring its accessibility for legitimate purposes poses a significant challenge. Conventional methods of protecting sensitive medical information have frequently proven inadequate, exposing vulnerabilities that jeopardize patient privacy. Blockchain technology has emerged as a promising solution due to its inherent characteristics of decentralization, immutability, and transparency. The objective of this paper is to propose a novel Blockchain-Assisted Optimized Privacy Preservation Model (BAOPPM) for Electronic Health Records (EHRs) that enhances the privacy of sensitive healthcare data while ensuring its accessibility for legitimate purposes. The model leverages blockchain technology for secure data storage and integrates advanced algorithms for efficient privacy key generation and data sanitization. At first, the HCK Algorithm is used to generate sensitive data clusters and non-sensitive data clusters. Then, the sensitive data are sanitized by adopting the Kronecker product between the sensitive data and the optimal key. Here, the optimal key is generated via the BMACO algorithm that considers the level of privacy by including privacy and privacy information ratio. Finally, the sanitized data are stored in the blockchain. On the contrary, the data restoration process is done, which is the reverse operation of data sanitization to retrieve the sensitive data from the blockchain. The BMACO approach achieves the lowest cost rate of 0.2451 at the 50th iteration, showcasing its superior convergence efficiency. In contrast, conventional methods such as COOT, BMO, KOA, COATI, LBO, EHO-OBL, and ECSO register higher cost ratings of 0.2463, 0.2498, 0.2512, 0.2484, 0.2473, 0.2536, and 0.2469, respectively. The sanitized data were stored in the blockchain and restored securely using the developed model.

Identifying brain tumors early and accurately is a significant way to improve patient outcomes, but access to numerous advanced diagnostic tools is not standardized across the world due to the cost and availability of MRI scans. We present a lightweight smartphone brain tumor diagnostic tool with deep learning–based diagnostic decision support in a contextualized way. We developed a convolutional neural network (CNN) based on MobileNetV2 for mobile deployment that allows for the processing of MRI images on consumer smartphones in real-time. The model was developed and validated on a publicly available brain tumor MRI dataset of glioma, meningioma, pituitary tumor, and normal cases, achieving an overall accuracy of 98% and classifying cases in less than 100 ms on standard iOS devices. This work demonstrates that with a lightweight architecture and on-device processing for the medical image, diagnostic decision support can be facilitated in a cost-effective, portable way, while also creating confidence factors in patient privacy, and represents an immense opportunity in lower-resourced clinical care, telemedicine, and point-of-care diagnosis around patients. It demonstrates another methodological option for the feasible implementation of advanced deep learning models to assist significant medical imaging workflows in a smartphone device.

This review article explores the application of artificial intelligence (AI) within Advanced Therapy Medicinal Products (ATMP) analysis, specifically focusing on challenges related to chemistry, manufacturing, and controls (CMC) and manufacturing processes. The inherent complexity and variability in ATMPs necessitate innovative solutions for potency testing, real-time process monitoring, and stability assessment. We examine how AI tools can contribute to these areas while navigating increasingly stringent regulatory landscapes. This work acknowledges the growing importance of data protection regulations worldwide, including frameworks such as HIPAA, GDPR, PIPEDA, POPIA, and LGPD, highlighting the need for secure data handling and patient privacy considerations within ATMP development and analysis. The integration of AI also necessitates attention to explainability and transparency, potentially leveraging techniques like SHAP values and physics-informed neural networks to ensure regulatory compliance and build trust in AI-driven insights.

The article examines the constitutional and legal regulation of medical secrecy disclosure in Ukraine, focusing on the balance between private and public interests. Medical secrecy is a cornerstone of human rights, ensuring privacy, personal inviolability, and personal data protection, while also playing a critical role in safeguarding national security, economic well-being, and fundamental rights. The study analyzes constitutional provisions, specialized legislation (e.g., Laws of Ukraine “On National Security of Ukraine” and “On Protection of the Population from Infectious Diseases”), and their application in exceptional cases where disclosure is deemed lawful, particularly for national security and economic stability. The evolution of Ukraine’s legislative approach reflects a shift from a broad, all-encompassing view of national security, including health care, to prioritizing sovereignty and democratic order, driven by wartime challenges. However, health care and access to medical information remain vital for economic and social stability. The article highlights cases where disclosure serves public interests, such as reporting occupational diseases, tracking infectious cases, and regulating during the COVID-19 pandemic, emphasizing adherence to legality, proportionality, and targeted data use. Additionally, disclosure is justified to protect human rights, including preventing infections among children, combating domestic violence, and ensuring vulnerable individuals’ safety. Ukraine’s legal framework is evolving toward a balanced model that harmonizes individual rights, societal needs, and state imperatives. This trajectory underscores the need for alignment with international human rights standards, enhancing resilience against global challenges like pandemics and digital threats. This balanced approach strengthens public trust in institutions, enhances citizen protection, and fosters a just, secure society for future generations.

Integrating biomedical information in healthcare organizations continues to be a challenge because of lack of cohesion, semantic variations, and tight privacy controls including the Health Insurance Portability and Accountability Act (HIPAA) and the General Data Protection Regulation (GDPR). Such limitations affect the collaboration analytics and hinder the move towards precision medicine. In this paper, we introduce OntoPharmX, a federated ontology-based semantic model, which can allow the integration of biomedical data with security, interoperability, and privacy through semantic networks. The suggested framework integrates semantic harmonization and HL7-FHIR adapters based on the OWL/RDF with cross-silo federated learning using secure aggregation and differential privacy to protect the privacy of patients without harming the analytical performance. OntoPharmX employs a standardized data transformation pipeline based on the modular pipeline of transforming ETL to FHIR to RDF and uses Flower/TensorFlow Federated to coordinate decentralized models and then an SPARQL-based reasoning. OntoPharmX has been experimentally validated across distributed healthcare locations to show that it can achieve near-centralized model accuracy and provide semantic alignment and can trade-off privacy and utility quantitatively. With the system, cohort discovery is privacy preserving and cross-institutional is achievable without undermining compliance or data sovereignty.

The right to privacy is an important value in medical ethics. This study was aimed at determining privacy-related attitudes and awareness of inpatients receiving health services in health institutions and revealing the affecting factors. In this descriptive, cross-sectional study, 194 patients hospitalized in the general surgery department of a university hospital were included. Data were collected using a two-part questionnaire. In the first part of the questionnaire, questions on the participant's descriptive characteristics and their knowledge about the subject were included. The second part included the "Patient Privacy Awareness and Attitude Scale." The mean age of the participants was 47.60±16.47years. Of them, 60.8% were women, 83.0% were knowledgeable about patient privacy and 90.2% had heard of the concept of patient rights. A significant relationship was determined between the participants' awareness of patient privacy, and variables such as age, education level, financial status and knowledge about privacy. In this study, the majority of the participants were knowledgeable of patient privacy. Almost all of them thought that healthcare personnel were careful about patient privacy. While medical interventions were implemented, it was observed that the participants' body privacy-related awareness levels were high. Patients' awareness of privacy is extremely important. Informational privacy is as important as physical privacy. Patients' awareness of privacy is extremely important. Informational privacy is as crucial as physical privacy. Patients' awareness of patient rights and privacy is crucial during the healthcare process. Patient rights are the embodiment of fundamental human rights within the context of the right to health. A heightened patient awareness of their rights is effective in increasing satisfaction with healthcare services. This study is believed to be effective in raising patient awareness of privacy. This study aimed to determine the privacy attitudes and influencing factors of patients in a surgical clinic. Privacy is an important value in medical practice.

Abstract—The Internet of Medical Things (IoMT) is rapidly transforming healthcare by enabling real-time monitoring, re- mote diagnosis, and intelligent decision-making. While these technologies improve patient care and efficiency, they also in- troduce new vulnerabilities in terms of data security, patient privacy, and system reliability. The growing reliance on inter- connected medical devices makes IoMT systems an attractive target for adversaries, with risks ranging from data breaches and adversarial manipulation to system-wide intrusions. Traditional security frameworks, such as centralized intrusion detection systems or rule-based approaches, struggle to keep up with the evolving nature of threats and the unique constraints of IoMT environments, including limited device resources, latency sensitivity, and the need for privacy preservation. To overcome these limitations, we present an integrated framework that combines federated learning, blockchain, and advanced deep learning models to provide a holistic solution for secure data processing and intrusion detection in IoMT ecosystems. The pro- posed architecture introduces quantum-based authentication for stronger device-level security, privacy-preserving collaborative training to enable distributed model learning without exposing raw patient data, and noise-driven feature masking to minimize the risks of adversarial attacks and poisoning attempts. In ad- dition, the framework reduces communication overhead through prototype-driven representation learning and optimization-aware aggregation, ensuring efficiency even in bandwidth-constrained medical networks. Index Terms—IoMT, Federated Learning, Blockchain, Deep Learning, Privacy Preservation, Intrusion Detection

Clinical registries are essential in oncology for monitoring the quality of patient care and supporting research. However, maintaining these registries is resource-intensive and can burden clinical staff. Technologies such as artificial intelligence (AI) now offer the ability to automatically extract data from electronic medical records into registries, with the potential to lower costs and improve efficiency. To examine the practical opportunities and challenges of automating oncology registries, using key lessons from the partial automation of the Australian Brain Cancer Registry (ABCR).The innovation:This analysis draws on the ABCR project experience, detailing the use of technologies ranging from discrete data extraction to advanced AI. It outlines the multidisciplinary approach required and discusses key factors relevant to registry automation.What can be learnt from this case?Successful registry automation relies on close collaboration between clinicians, researchers and programmers. Human oversight remains essential, particularly when the AI is uncertain about specific data points. Key factors for effective automation include clearly defined data elements, strong communication among stakeholders, robust safeguards for patient privacy and planning for long-term sustainability and interoperability of the registry. It is also important to avoid introducing bias by over-prioritising data that are easiest to extract automatically. Automating cancer registries can reduce costs but requires thorough planning. The optimal approach may involve humans and machines working together.Implications for health information management practice:Giving early attention to data accuracy, patient privacy and the long-term sustainability of the registry is critical for long-term success.

Background: Hypertrophic cardiomyopathy (HCM) is the most common heritable cardiac disease, yet it remains profoundly underdiagnosed; it is estimated that 6 out of 7 (86%) individuals with HCM have not been identified and are unaware of their diagnosis. HCM is typically diagnosed using routine echocardiography. However, the cost, time to perform each study, need for patient privacy, and inherent issues with transporting large echocardiogram machines, has limited the use of echocardiography for wide-scale community-based screening. Hypothesis: Given that the parasternal long axis (PLAx) view can visualize many of the characteristic findings seen in HCM, we hypothesized that a streamlined, cost-effective approach to the diagnosis of HCM would utilize a point-of-care echo probe for obtaining a single PLAx sweep of the heart (SPLASH) + color Doppler to screen for HCM. Here we report the sensitivity and specificity of our SPLASH screening. Methods: We studied 108 consecutive patients presenting to our echocardiography laboratory for exercise stress testing on days that HCM patients and patients with chronic kidney disease (CKD) and end stage renal disease are routinely screened. SPLASH echocardiograms were performed by an experienced sonographer. Each was read independently in random order by 2 echo attendings and 2 independent sonographers, all blinded to patient diagnosis. Results were categorized as definite HCM, probable HCM, possible HCM, probably not HCM, and no evidence of HCM. Readers also indicated whether obtaining additional views for further assessment was indicated. Results were compared to the gold standard of full echocardiographic result combined with clinical diagnosis. Results: The mean age of the cohort was 57.3 years (53% women; 22% self-reported as Black). 63% had hypertension and 35% had CKD. Time to perform one SPLASH study was, on average, less than 5 minutes. Specificity of the SPLASH was 91.0% for attendings and 83.7% for sonographers. Sensitivity was 84.5% for attendings and 85% for sonographers. When the studies that the reader indicated needed additional views for further assessment were included, the sensitivity increased to 92.5%. Conclusions: We report a novel and streamlined method for performing echocardiographic screening for HCM with high sensitivity and specificity. If validated in community settings, this method has the potential to change medical practice and improve access to rapid yet accurate diagnostic testing for HCM.

The nnU-Net framework has played a crucial role in medical image segmentation and has become the gold standard in multitudes of applications targeting different diseases, organs, and modalities. However, so far it has been used primarily in a centralized approach where the collected data is stored in the same location where nnU-Net is trained. This centralized approach has various limitations, such as potential leakage of sensitive patient information and violation of patient privacy. Federated learning has emerged as a key approach for training segmentation models in a decentralized manner, enabling collaborative development while prioritising patient privacy. In this paper, we propose FednnU-Net, a plug-and-play, federated learning extension of the nnU-Net framework. To this end, we contribute two federated methodologies to unlock decentralized training of nnU-Net, namely, Federated Fingerprint Extraction (FFE) and Asymmetric Federated Averaging (AsymFedAvg). We conduct a comprehensive set of experiments demonstrating high and consistent performance of our methods for breast, cardiac and fetal segmentation based on a multi-modal collection of 6 datasets representing samples from 18 different institutions. To democratize research as well as real-world deployments of decentralized training in clinical centres, we publicly share our framework at https://github.com/faildeny/FednnUNet.

Artificial intelligence (AI) is in the era of rapid evolution. Like other healthcare fields, AI has significantly impacted sleep medicine. We aim to explain the evolving role of AI in sleep medicine and provide clinicians with key information related to its benefits and limitations. AI technologies, like machine learning and deep learning, improve the detection of sleep disorders, such as obstructive sleep apnea, insomnia, and narcolepsy, through advanced data analysis from tools like polysomnography and consumer sleep devices. AI also enables targeted therapies by endotyping sleep disorders, optimizing patient care, and reducing unnecessary treatments. On the other hand, there are many challenges that need to be addressed before using AI in clinical settings. Ethical issues regarding patient privacy, biases, and transparency regarding data use are some of the key challenges. AI could transform sleep medicine by enhancing diagnostic accuracy and personalizing treatment plans. Effective collaboration between clinicians and AI experts is necessary to use AI optimally in clinical settings.

Patient Confidentiality and Privacy Among Jordanian Medical Students: Factors Associated with Knowledge and Practice