Gut microbiome research has made tremendous progress, especially with the integration of machine learning and artificial intelligence that can provide new insights from complex microbiome data and its impact on human health. The use of explainable artificial intelligence is becoming critical in medicine and adopting it in precision medicine-models leveraging gut microbiome data is appealing for providing more transparency and trustworthiness in clinical research. This scoping review evaluates the use of machine learning and explainable artificial intelligence techniques and identifies existing gaps in knowledge in this research area to suggest future research directions. Online databases (PubMed and Scopus) were searched to retrieve papers published between 2018-2024, and from which we selected 76 publications. Different clinical applications of machine learning and artificial intelligence techniques in gut microbiome studies were explored in the reviewed articles. We observed a high prevalence in the use of black box models in the field, with Random Forest being the most used algorithm. The explainability remains somewhat limited in the field, but it appears to be improving. Researchers showed interest in SHAP applications as an explainable technique. Finally, not enough attention was paid to the reproducibility of the research work published. This review highlights opportunities for advancing research on explainable artificial intelligence models in the field of microbiome, supporting future applications of microbiome-based precision medicine.

The European Health Data Space (EHDS) is a significant initiative within the European Union aimed at improving healthcare through better use of health data and collaboration. This involves sharing and utilization of health data in a secure and interoperable manner across EU member states. With the advent of modern technology, different types of health data are being acquired ranging from traditional electronic health records to real-time wearable signals for improved healthcare delivery and management. However, there are certain challenges associated with compliance to EHDS, such as secure continuous integration (CI) and continuous delivery (CD), scalability and adaptability towards new technologies and applications, tolerance against IoT cyber-attacks, requirement of advanced data storage methods, processing power, interoperability solutions for data analysis and connectivity. In order to address these, a flagship project of the European Commission, GATEKEEPER, represents a pioneering effort designed to transform healthcare delivery through the utilization of sophisticated digital technologies. The architecture of the GATEKEEPER platform has been meticulously crafted, in compliance with the EHDS, to integrate a multitude of data sources, deploying advanced analytical tools to enable the development of novel solutions for the effective prediction, detection, and management of chronic health conditions. GATEKEEPER addresses the aforementioned challenges by integrating collaborative tools for data management and consistency, machine learning modelling and management tool, activity feeds and issue tracking tools with developers authentication, CI/CD tools allowing, among others scalability in multiple formats, tools for assessing security, vulnerability and dependency for third party packages.

Artificial intelligence and machine learning (AI/ML) are playing increasingly important roles, permeating the field of medical devices (MDs). This rapid progress has not yet been matched by the Health Technology Assessment (HTA) process, which still needs to define a common methodology for assessing AI/ML-based MDs. To collect existing evidence from the literature about the methods used to assess AI-based MDs, with a specific focus on those used for the management of heart failure (HF), the International Federation of Medical and Biological Engineering (IFMBE) conducted a scoping meta-review. This manuscript presents the results of this search, which covered the period from January 1974 to October 2022. After careful independent screening, 21 reviews, mainly conducted in North America and Europe, were retained and included. Among the findings were that deep learning is the most commonly utilised method and that electronic health records and registries are among the most prevalent sources of data for AI/ML algorithms. Out of the 21 included reviews, 19 focused on risk prediction and/or the early diagnosis of HF. Furthermore, 10 reviews provided evidence of the impact on the incidence/progression of HF, and 13 on the length of stay. From an HTA perspective, the main areas requiring improvement are the quality assessment of studies on AI/ML (included in 11 out of 21 reviews) and their data sources, as well as the definition of the criteria used to assess the selection of the most appropriate AI/ML algorithm.

Purpose: This study evaluates the effects of non-invasive home telemonitoring methods of managing congestive heart failure (CHF) patients with particular focus on complexity of intervention, patient characteristics, patient severity, and key enabling technologies (KETs) adopted. Our goal was to capture all possible aspects of previously documented outcomes and provide updated and clearer evidence on mixed effects on common themes. Methods: Randomized controlled trials (RCTs) published between 1 January 2012 and 6 June 2019, focusing on home telemonitoring of patients with only CHF or CHF coexisting with other chronic disease(s), were retrieved from online resources (PubMed, Embase, MEDLINE(R), Your journals@Ovid, Elsevier, and ClinicalTrials.gov). The snowball sampling method and forward citation tracking on Google Scholar were also adopted to identify additional relevant studies. Retrieved studies were in a language known by the authors (i.e., English, Spanish or Italian). Quality assessment of individual studies for shortcomings in design, management, evaluation, and reporting was done using the Cochrane risk of bias (RoB) tool. Variables of interest were synthesized as differences in relative risk (RR), or as weighted mean differences (WMD). Outcomes were assigned as primary or secondary based on a principal judgement of clinical importance, and secondarily on highest recurrent counts in included studies. Results: In all, 28 RCTs involving 10,258 patients were included in the qualitative synthesis, out of which 24 were used for the quantitative synthesis. These studies focused on non-invasive telemonitoring practices for home monitoring of CHF patients, through the deployment of different kinds of electronic/mobile devices, with most having wireless communication capabilities. Moreover, studies focusing on implantable monitoring devices in terms of inputs, data and patient performance were also included. Brain natriuretic peptide (BNP) ((WMD = -27.75; 95% CI (-53.36, -2.14); p-value = 0.034), rehospitalization/hospitalization for heart failure (RR = 0.88; 95% CI (0.79, 0.98); p-value = 0.015), cardiovascular death/heart failure hospitalization (RR = 0.70; 95% CI (0.51, 0.97); p-value = 0.03), and six minute walk test (6MWT) (WMD = 25.61; 95% CI (9.22, 41.99); p-value = 0.002) significantly improved in the telemonitoring group, while the number of visits to a nurse (WMD = 1.42; 95% CI (0.33, 2.52); p-value = 0.011) increased considerably compared to usual care. Conclusion: Although there were limitations to the evidence provided in this review such as wide variations in certain variables (e.g., sample populations, RoB assessment, telemonitoring tools, follow-up periods), issues with allocation concealment and blinding of participants and personnel, and paucity of data for synthesizing particular outcomes of interest, overall, telemonitoring seems to offer much better results in the treatment of CHF patients compared to usual care. Registration and protocol: This systematic review and meta-analysis has been retrospectively registered in the Open Science Framework (OSF) repository with https://doi.org/10.17605/OSF.IO/NDXCP. All data related to this study, including the electronic supplementary data, can be found at this link: osf.io/57q3h.

Monitoring of advanced cancer patients' health, treatment, and supportive care is essential for improving cancer survival outcomes. Traditionally, oncology has relied on clinical metrics such as survival rates, time to disease progression, and clinician-assessed toxicities. In recent years, patient-reported outcome measures (PROMs) have provided a complementary perspective, offering insights into patients' health-related quality of life (HRQoL). However, collecting PROMs consistently requires frequent clinical assessments, creating important logistical challenges. Wearable devices combined with artificial intelligence (AI) present an innovative solution for continuous, real-time HRQoL monitoring. While deep learning models effectively capture temporal patterns in physiological data, most existing approaches are unimodal, limiting their ability to address patient heterogeneity and complexity. This study introduces a multimodal deep learning approach to estimate HRQoL in advanced cancer patients. Physiological data, such as heart rate and sleep quality collected via wearable devices, are analyzed using a hybrid model combining convolutional neural networks (CNNs) and bidirectional long short-term memory (BiLSTM) networks with an attention mechanism. The BiLSTM extracts temporal dynamics, while the attention mechanism highlights key features, and CNNs detect localized patterns. PROMs, including the Hospital Anxiety and Depression Scale (HADS) and the Integrated Palliative Care Outcome Scale (IPOS), are processed through a parallel neural network before being integrated into the physiological data pipeline. The proposed model was validated with data from 204 patients over 42 days, achieving a mean absolute percentage error (MAPE) of 0.24 in HRQoL prediction. These results demonstrate the potential of combining wearable data and PROMs to improve advanced cancer care.