Artificial Intelligence for Elderly Fall Detection: State-of-the-art Methods, Applications and Challenges

Continual learning in medical devices: FDA's action plan and beyond

Chemical heuristics have been fundamental to the advancement of chemistry and materials science. These heuristics are typically established by scientists using knowledge and creativity to extract patterns from limited datasets. Machine learning offers opportunities to perfect this approach using computers and larger datasets. Here, we discuss the relationships between traditional heuristics and machine learning approaches. We show how traditional rules can be challenged by large-scale statistical assessment and how traditional concepts commonly used as features are feeding the machine learning techniques. We stress the waste involved in relearning chemical rules and the challenges in terms of data size requirements for purely data-driven approaches. Our view is that heuristic and machine learning approaches are at their best when they work together.

Artificial Intelligence: The Future of Maxillofacial Prognosis and Diagnosis?

Artificial intelligence (AI) has, in the recent past, experienced a rebirth with the growth of generative AI systems such as ChatGPT and Bard. These systems are trained with billions of parameters and have enabled widespread accessibility and understanding of AI among different user groups. Widespread adoption of AI has led to the need for understanding how machine learning (ML) models operate to build trust in them. An understanding of how these models generate their results remains a huge challenge that explainable AI seeks to solve. Federated learning (FL) grew out of the need to have privacy-preserving AI by having ML models that are decentralized but still share model parameters with a global model. This study sought to examine the extent of development of the explainable AI field within the FL environment in relation to the main contributions made, the types of FL, the sectors it is applied to, the models used, the methods applied by each study, and the databases from which sources are obtained. A systematic search in 8 electronic databases, namely, Web of Science Core Collection, Scopus, PubMed, ACM Digital Library, IEEE Xplore, Mendeley, BASE, and Google Scholar, was undertaken. A review of 26 studies revealed that research on explainable FL is steadily growing despite being concentrated in Europe and Asia. The key determinants of FL use were data privacy and limited training data. Horizontal FL remains the preferred approach for federated ML, whereas post hoc explainability techniques were preferred. There is potential for development of novel approaches and improvement of existing approaches in the explainable FL field, especially for critical areas. OSF Registries 10.17605/OSF.IO/Y85WA; https://osf.io/y85wa.

The aging and multimorbid population and health personnel shortages pose a substantial burden on primary health care. While predictive machine learning (ML) algorithms have the potential to address these challenges, concerns include transparency and insufficient reporting of model validation and effectiveness of the implementation in the clinical workflow. To systematically identify predictive ML algorithms implemented in primary care from peer-reviewed literature and US Food and Drug Administration (FDA) and Conformité Européene (CE) registration databases and to ascertain the public availability of evidence, including peer-reviewed literature, gray literature, and technical reports across the artificial intelligence (AI) life cycle. PubMed, Embase, Web of Science, Cochrane Library, Emcare, Academic Search Premier, IEEE Xplore, ACM Digital Library, MathSciNet, AAAI.org (Association for the Advancement of Artificial Intelligence), arXiv, Epistemonikos, PsycINFO, and Google Scholar were searched for studies published between January 2000 and July 2023, with search terms that were related to AI, primary care, and implementation. The search extended to CE-marked or FDA-approved predictive ML algorithms obtained from relevant registration databases. Three reviewers gathered subsequent evidence involving strategies such as product searches, exploration of references, manufacturer website visits, and direct inquiries to authors and product owners. The extent to which the evidence for each predictive ML algorithm aligned with the Dutch AI predictive algorithm (AIPA) guideline requirements was assessed per AI life cycle phase, producing evidence availability scores. The systematic search identified 43 predictive ML algorithms, of which 25 were commercially available and CE-marked or FDA-approved. The predictive ML algorithms spanned multiple clinical domains, but most (27 [63%]) focused on cardiovascular diseases and diabetes. Most (35 [81%]) were published within the past 5 years. The availability of evidence varied across different phases of the predictive ML algorithm life cycle, with evidence being reported the least for phase 1 (preparation) and phase 5 (impact assessment) (19% and 30%, respectively). Twelve (28%) predictive ML algorithms achieved approximately half of their maximum individual evidence availability score. Overall, predictive ML algorithms from peer-reviewed literature showed higher evidence availability compared with those from FDA-approved or CE-marked databases (45% vs 29%). The findings indicate an urgent need to improve the availability of evidence regarding the predictive ML algorithms' quality criteria. Adopting the Dutch AIPA guideline could facilitate transparent and consistent reporting of the quality criteria that could foster trust among end users and facilitating large-scale implementation.

Artificial Intelligence (AI) and Machine Learning (ML) are becoming common in our daily lives. The AI-driven processes significantly affect us as individuals and as a society, spanning across ethical dimensions like discrimination, misinformation, and fraud. Several of these AI & ML approaches rely on Knowledge Graph (KG) data. Due to the large volume and complexity of today's KG-driven approaches, enormous resources are spent to utilize the complex AI approaches. Efficient usage of the resources like hardware and power consumption is essential for sustainable KG-based ML technologies. This paper introduces the ethical and sustainability considerations, challenges, and optimizations in the context of KG-based ML. We have grouped the ethical and sustainability aspects according to the typical Research & Development (R&D) lifecycle: an initial investigation of the AI approach's responsibility dimensions; technical system setup; central KG data analytics and curating; model selection, training, and evaluation; and final technology deployment. We also describe significant trade-offs and alternative options for dedicated scenarios enriched through existing and reported ethical and sustainability issues in AI-driven approaches and research. These include, e.g., efficient hardware usage guidelines; or the trade-off between transparency and accessibility compared to the risk of manipulability and privacy-related data disclosure. In addition, we propose how biased data and barely explainable AI can result in discriminating ML predictions. This work supports researchers and developers in reflecting, evaluating, and optimizing dedicated KG-based ML approaches in the dimensions of ethics and sustainability.

An awakening in medicine: the partnership of humanity and intelligent machines

Artificial intelligence and Machine Learning approaches in sports: Concepts, applications, challenges, and future perspectives

The advancement in technologies such as Artificial Intelligence (AI) and the Internet of Things (IoT) has enabled the development of a variety of healthcare applications and services such as Remote Health Monitoring (RHM). Developing an RHM system requires integrating various medical sensors that generate a vast amount of data. The integration of AI and Machine Learning (ML) models in the RHM system for analysis of the medical data can enable the development of a decision support system that can reduce costs and save valuable time for clinicians and caregivers. This thesis is focused on leveraging ML and AI technique sto address the challenges in RHM with a specific emphasis on improving the performance of AI-based decision support systems. This is achieved by addressing limitations in current AI-based models and developing novel Random Neural Network (RNN)-based models for fall detection and epilepsy detection and prediction. Initially the thesis presents a literature review on state-of-the-art machine learning techniques used in remote healthcare solutions, automatic fall activity recognition for elderly populations, and epileptic seizure detection and prediction. Furthermore, the study presents a novel fall detection method based on RNN. The results obtained from the proposed model are compared with traditional classification algorithms such as Artificial Neural Networks, Convolutional Neural Networks, and state-of-the-art fall detection schemes. The proposed RNN-based method demonstrates significant improvements in accuracy compared to other methods. The proposed scheme achieved an overall accuracy of 98% in detecting falls. Additionally, several other parameters such as precision, recall, specificity, and F-measure show that the proposed algorithm has better generalization capabilities when compared with other traditional machine learning schemes. Furthermore, the proposed RNN is also compared with a recent scheme and the obtained results demonstrate the superiority of the proposed scheme. Additionally, the thesis presents two novel RNN-based classification schemes for detecting epileptic seizures. In the first scheme statistical features are extracted from raw EEG data, while the second method incorporates frequency analysis techniques such as Discrete Wavelet Transform (DWT) and Fast Fourier Transform. The performance of these models is evaluated and compared with other classification algorithms, demonstrating the effectiveness of RNN-based models for epilepsy detection. Finally, a novel RNN-based epilepsy prediction model is proposed for classification of EEG data into three classes i.e. ictal, pre-ictal, and inter-ictal. The model pre-processes the EEG data by decomposing the signals into five frequency sub-bands using DWT. After pre-processing, six statistical features are extracted from each frequency sub-band which a reused as an input vector to train the RNN model. The proposed scheme achieves an overall accuracy of 95.66% in seizure prediction, outperforming other state-of-the-art models.

In the context of escalating global polycrises, residential high-rises face unprecedented challenges that threaten infrastructure and occupant safety. This literature review examines the role of Artificial Intelligence (AI) in managing these crises within vertical communities. Covering studies from 2014 to 2024, the review utilized a comprehensive search strategy across Web of Science, Scopus, IEEE Xplore, Google Scholar, and PubMed, focusing on AI, polycrisis, and high-rise buildings. The analysis identified 87 relevant studies showcasing the potential of AI technologies — such as machine learning, natural language processing, and computer vision — to enhance crisis management. Key findings indicate that AI can significantly improve emergency preparedness, real-time information dissemination, and resource optimization. Computer vision advances hazard detection accuracy, facilitating faster evacuations and better safety. The integration of IoT sensor networks with predictive analytics reduces false alarms and enhances early threat detection. However, the review also highlights challenges, including data privacy, cybersecurity, scalability, and the balance between AI autonomy and human oversight. Future research should address these limitations by focusing on real-world case studies to assess AI performance across diverse crisis scenarios. Development of more robust AI models and their integration with existing crisis management frameworks is essential. Ethical considerations, particularly regarding privacy and bias, must also be scrutinized to ensure that AI solutions are both effective and equitable. This review provides valuable insights into using AI to bolster urban resilience and support Sustainable Development Goals in high-rise residential settings, emphasizing the need for strategic AI integration and ongoing research to address current gaps and challenges.

Theory-Guided Machine Learning Finds Geometric Structure-Property Relationships for Chemisorption on Subsurface Alloys

The primary objective of this study is to investigate various applications of artificial intelligence (AI) and statistical methodologies for analyzing and managing peritoneal metastases (PM) caused by gastrointestinal cancers. Relevant keywords and search criteria were comprehensively researched on PubMed and Google Scholar to identify articles and reviews related to the topic. The AI approaches considered were conventional machine learning (ML) and deep learning (DL) models, and the relevant statistical approaches included biostatistics and logistic models. The systematic literature review yielded nearly 30 articles meeting the predefined criteria. Analyses of these studies showed that AI methodologies consistently outperformed traditional statistical approaches. In the AI approaches, DL consistently produced the most precise results, while classical ML demonstrated varied performance but maintained high predictive accuracy. The sample size was the recurring factor that increased the accuracy of the predictions for models of the same type. AI and statistical approaches can detect PM developing among patients with gastrointestinal cancers. Therefore, if clinicians integrated these approaches into diagnostics and prognostics, they could better analyze and manage PM, enhancing clinical decision-making and patients' outcomes. Collaboration across multiple institutions would also help in standardizing methods for data collection and allowing consistent results.

Diabetes is a lifelong condition which is associated with abnormally high blood sugar levels due to insufficient production of insulin or failure of the body to use the hormone efficiently. Due to a rising number of diabetes cases globally, the need to develop efficient and targeted approaches for the disease is more pressing than pre/application. Artificial intelligence (AI) and machine learning (ML) are among the most significant innovations in the healthcare industry, creating new positive directions for creating individualized plans for diabetes mellitus management. AI and ML are not just limited to handling medical and patients’ data, but it helps in robot assisted surgeries, virtual nursing professionals, and computer aided diagnosis. These technologies are not deploying human practitioners out of service, but instead heightening their capabilities culminating in a health win and decrease in cost. PROAC supports further research and development of these fast-growing technologies in healthcare, as we find ourselves on the brink of revolution that might change disease management and customization of treatment for many more years to come. This review has therefore involved using different electronic databases, including PubMed, Scopus and Google Scholar, to search for relevant published literature. The application of AI and ML has proved promising, and the research concluded that it can help develop a risk assessment model for early diagnosis and prevention of diabetes. Such algorithms look at a combination of multiple risk factors including family history, genetics, nutrition, biochemical markers, and physical activity level, which puts those with higher risk on special diet, exercise regimens or screening. Further, AI endorsing, decision support systems may provide tailored recommendations regarding treatment schedules based on patients’ characteristics such as insulin response, diet and exercise. Additionally, the continuous glucose monitoring (CGM) devices integrable with the machine learning algorithms can inform real-time information about glucose behaviors to make relevant alterations to doses of insulin and a dietary intake. This paper presents various benefits of adopting AI and ML in diabetes care: heterogeneous and accurate risk prediction, individualized therapies, and observation. But the problem of data quality, data privacy issues and the issue of interpretability and explainability of the trained AI models remains a problem in the use of AI in automated scheduling. The approaches require interdisciplinary collaborations across the medical field, data analytics, and policy departments in order to be properly applied and used.

Introduction. The intensive care unit (ICU) plays a pivotal role in providing specialized care to patients with severe illnesses or injuries. As a critical aspect of healthcare, ICU admissions demand immediate attention and skilled care from healthcare professionals. However, the intricacies involved in this process necessitate analytical solutions to ensure effective management and optimal patient outcomes. Aim. The aim of this review was to highlight the enhancement of the ICUs through the application of analytics, artificial intelligence, and machine learning. Methods. The review approach was carried out through databases such as MEDLINE, Embase, Web of Science, Scopus, Taylor & Francis, Sage, ProQuest, Science Direct, CINAHL, and Google Scholar. These databases were chosen due to their potential to offer pertinent and comprehensive coverage of the topic while reducing the likelihood of overlooking certain publications. The studies for this review involved the period from 2016 to 2023. Results. Artificial intelligence and machine learning have been instrumental in benchmarking and identifying effective practices to enhance ICU care. These advanced technologies have demonstrated significant improvements in various aspects. Conclusions. Artificial intelligence, machine learning, and data analysis techniques significantly improved critical care, patient outcomes, and healthcare delivery.

Background: Artificial intelligence (AI) has emerged as a transformative tool in modern medicine, particularly in dermatology, where it supports the diagnosis and management of various skin diseases, including skin cancer. Through machine learning and deep learning techniques, AI enables accurate analysis of clinical and dermoscopic images, improving early detection and clinical outcomes. Objective: This systematic review aimed to evaluate the clinical applications of AI in dermatology, focusing on its impact on diagnostic accuracy, workflow efficiency, and access to specialized care. Methods: The review was conducted according to PRISMA guidelines. Peer-reviewed studies published between January 2020 and March 2025 in English or Spanish were included if they evaluated AI-based tools for dermatological diagnosis, classification, or treatment. Animal studies, editorials, non-peer-reviewed articles, and studies with an unclear methodology were excluded. A comprehensive search was performed in PubMed, Scopus, IEEE Xplore, and Google Scholar between December 2024 and March 2025. The risk of bias was assessed qualitatively, using a tailored framework based on study design, dataset transparency, and clinical applicability. Results: A total of 29 studies met the inclusion criteria. AI tools demonstrated high performance in melanoma detection, achieving up to 90% accuracy and 85% sensitivity. In clinical settings, AI support reduced mismanagement of malignant lesions from 58.8% to 4.1% and avoided 27% of unnecessary procedures in benign cases. Additional tools such as convolutional neural networks and imaging systems like FotoFinder also showed promising results. Limitations: Limitations of the evidence include the heterogeneity of AI models, lack of external validation, and a moderate-to-high risk of bias. Conclusions: AI has demonstrated robust clinical potential in dermatology, particularly in cancer detection and workflow optimization. However, further studies are required to address challenges such as algorithmic bias, data privacy, and regulatory oversight. Funding and registration: This review received no external funding and was not registered in a systematic review registry.