The adaptive nature of artificial intelligence (AI), with its ability to improve performance through continuous learning, offers substantial benefits across various sectors. However, current regulatory frameworks are not intended to accommodate this adaptive nature, and they have prolonged approval timelines, sometimes exceeding one year for some AI-enabled devices. This creates significant challenges for manufacturers who must deal with lengthy waits and submit multiple approval requests for AI-enabled device software functions as they are updated. In response, regulatory agencies like the US Food and Drug Administration (FDA) have introduced guidelines to better support the approval process for continuously evolving AI technologies. This article explores the FDA’s concept of predetermined change control plans and how they can streamline regulatory oversight by reducing the need for repeated approvals, while ensuring safety and compliance. This can help reduce the burden for regulatory bodies and decrease waiting times for approval decisions, therefore fostering innovation, increasing market uptake, and exploiting the benefits of artificial intelligence and machine learning technologies.

Background: Retinal diseases are emerging as a significant health concern, making early detection and prompt treatment crucial to prevent visual impairment. Optical coherence tomography (OCT) is the preferred imaging modality for non-invasive diagnosis. Both diabetic macular edema (DME) and macular edema secondary to retinal vein occlusion (RVO) present an increase in retinal thickness, posing etiologic diagnostic challenges for non-specialists in retinal diseases. The lack of research on deep learning classification of macular edema secondary to RVO using OCT images motivated us to propose a convolutional neural network model for this task. Methods: The VGG-19 network was fine-tuned with a public dataset to classify OCT images. This network was then used to develop three models: unimodal—the input is only the OCT B-scan; multimodal—the inputs are the OCT B-scan and diabetes information, and multi-image—the inputs are the OCT B-scan, the infrared image, and the diabetes information. Seven hundred sixty-six patients from ULS São João were selected, comprising 208 healthy eyes, 207 with macular edema secondary to RVO, 218 with DME, and 200 with other pathologies. The performance metrics include accuracy, precision, recall, F_0.5 score, and area under the receiver operating characteristic curve (AUROC). Results: The multi-image model achieved better results, with an accuracy of 95.20%, precision of 95.43%, recall of 95.20%, F_0.5-score of 95.32%, F₁-score of 95.21%, and AUROC of 99.59% on the classification task between four classes. Conclusions: This study presents a novel method to distinguish macular edema secondary to RVO and DME using diabetes diagnosis, OCT, and infrared images. This research aims to provide a reliable tool for ophthalmologists, improving the accuracy and speed of diagnoses.

Background/Objectives: Colorectal anastomotic leak (CAL) is one of the most severe postoperative complications in colorectal surgery, impacting patient morbidity and mortality. Current risk assessment methods rely on clinical and intraoperative factors, but no real-time predictive tool exists. This study aimed to develop an artificial intelligence model based on intraoperative laparoscopic recording of the anastomosis for CAL prediction. Methods: A convolutional neural network (CNN) was trained with annotated frames from colorectal surgery videos across three international high-volume centers (Instituto Português de Oncologia de Lisboa, Hospital das Clínicas de Ribeirão Preto, and Royal Liverpool University Hospital). The dataset included a total of 5356 frames from 26 patients, 2007 with CAL and 3349 showing normal anastomosis. Four CNN architectures (EfficientNetB0, EfficientNetB7, ResNet50, and MobileNetV2) were tested. The models’ performance was evaluated using their sensitivity, specificity, accuracy, and area under the receiver operating characteristic (AUROC) curve. Heatmaps were generated to identify key image regions influencing predictions. Results: The best-performing model achieved an accuracy of 99.6%, AUROC of 99.6%, sensitivity of 99.2%, specificity of 100.0%, PPV of 100.0%, and NPV of 98.9%. The model reliably identified CAL-positive frames and provided visual explanations through heatmaps. Conclusions: To our knowledge, this is the first AI model developed to predict CAL using intraoperative video analysis. Its accuracy suggests the potential to redefine surgical decision-making by providing real-time risk assessment. Further refinement with a larger dataset and diverse surgical techniques could enable intraoperative interventions to prevent CAL before it occurs, marking a paradigm shift in colorectal surgery.

Introduction: The European Union prioritizes occupational health and safety, with firefighters—particularly volunteers—facing significant risks. In Portugal, 94% of firefighters are volunteers, often balancing multiple jobs, which increases their exposure to hazards. This study investigates Portuguese firefighters’ occupational risks and identifies tasks with the highest accident incidence to build a comprehensive profile of their health and work conditions. Background: Firefighters are exposed to extreme environments, long shifts, and high-stress situations, especially during wildfires. These conditions contribute to physical and mental health issues, including cardiovascular disease and sleep disturbances. Training and experience are key to improving risk perception and preventing accidents. Understanding firefighters’ main characteristics is essential for risk reduction and improving their quality of life. Methods: A sociodemographic questionnaire was developed and administered to firefighters, covering three sections: General Information, Lifestyle and Clinical History, and Firefighting Activity. This enabled the collection of sociodemographic and occupational data. Results: A total of 220 responses were collected from Portuguese volunteer firefighters. Cardiovascular and respiratory diseases were the most commonly reported diagnoses. Wildland and urban/industrial fires were the main operational contexts linked to accidents, with 84 individuals reporting accident history in these scenarios. Discussion/Conclusion: The study revealed a lack of data on the profile of volunteer firefighters, a group at heightened risk due to cumulative exposures and existing health conditions. Further research is needed to support the development of regulations and health monitoring strategies that protect this workforce without compromising their operational capacity.

Opaque deep learning models in biomedical applications pose challenges related to bias, fairness, and regulatory compliance. This paper presents a narrative review of explainable artificial intelligence (XAI) in healthcare, specifically medical diagnosis, based on a targeted selection guided by a narrative review of recent literature, from 2019 to 2026, conducted on Scopus, Web of Science, and PubMed databases. Covered topics comprise model interpretability, including image-based models and multimodal models, with special focus on large language models (LLMs) as a hard-to-interpret system. The review situates XAI within key regulatory frameworks, including the European Union (EU) Artificial Intelligence (AI) Act, highlighting how explainability supports legal requirements for transparency, auditability, and accountability. It examines widely used XAI methods such as Shapley Additive Explanations (SHAP), Local Interpretable Model-Agnostic Explanations (LIME), attention visualization, neurosymbolic reasoning, and others. These techniques help identify bias, detect spurious correlations, and analyze hallucinations in LLM deployment. They present an effort to, within a feasible range, minimize these events and increase confidence in the predictions provided, without dispensing with clear and proper validation within the medical classification loop. While emphasizing the benefits of XAI, the review also addresses key limitations, including limited explanation fidelity, cognitive overload, and the risk of misleading interpretations. Overall, it provides a concise synthesis of how explainability intersects with ethics, regulation, and system design to support oversight and transparency in AI healthcare. Additionally, it provides suggestions for the incorporation of technical XAI methods within current legislation and guidelines to provide a robust framework for AI model deployment.

Background: Retinal diseases are becoming an important public health issue, with early diagnosis and timely intervention playing a key role in preventing vision loss. Optical coherence tomography (OCT) remains the leading non-invasive imaging technique for identifying retinal conditions. However, distinguishing between diabetic macular edema (DME) and macular edema resulting from retinal vein occlusion (RVO) can be particularly challenging, especially for clinicians without specialized training in retinal disorders, as both conditions manifest through increased retinal thickness. Due to the limited research exploring the application of deep learning methods, particularly for RVO detection using OCT scans, this study proposes a novel diagnostic approach based on stacked convolutional neural networks. This architecture aims to enhance classification accuracy by integrating multiple neural network layers, enabling more robust feature extraction and improved differentiation between retinal pathologies. Methods: The VGG-16, VGG-19, and ResNet50 models were fine-tuned using the Kermany dataset to classify the OCT images and afterwards were trained using a private OCT dataset. Four stacked models were then developed using these models: a model using the VGG-16 and VGG-19 networks, a model using the VGG-16 and ResNet50 networks, a model using the VGG-19 and ResNet50 models, and finally a model using all three networks. The performance metrics of the model includes accuracy, precision, recall, $F_2$-score, and area under of the receiver operating characteristic curve (AUROC). Results: The stacked neural network using all three models achieved the best results, having an accuracy of 90.7%, precision of 99.2%, a recall of 90.7%, and an $F_2$-score of 92.3%. Conclusions: This study presents a novel method for distinguishing retinal disease by using stacked neural networks. This research aims to provide a reliable tool for ophthalmologists to improve diagnosis accuracy and speed.