❌

Reading view

↗️

Integrating liquid biopsies and artificial intelligence for early cancer detection: A systematic review and meta-analysis

✇(Multiomics OR Omics) AND (Pancreatic)
By: Panagiotis Filis · Georgios Markozannes · Dimitrios Salgkamis · Nikos Tsiknakis · Ioannis Zerdes · Eirini Pagkalidou · Maria Manou · Nikolaos Filis · Andri Papakonstantinou · Dimitrios Mavroudis · Alexios Matikas · Konstantinos K Tsilidis · Theodoros Foukakis

Eur J Cancer. 2026 Mar 24;239:116699. doi: 10.1016/j.ejca.2026.116699. Online ahead of print.

ABSTRACT

INTRODUCTION: The latest generation of liquid biopsies incorporates multi-omic features, including genomics, methylomics, and fragmentomics. Machine learning (ML) approaches have been proposed to synthesize these complex biological data for the development of diagnostic classifiers. This study aims to evaluate the integration of ML with circulating cell-free DNA (cfDNA) analysis for early cancer detection.

METHODS: Medline, Embase, Cochrane, and Web of Science were searched in July 2025. Eligible studies combined ML and cfDNA features to distinguish cancer patients (stages I-III) from non-cancer controls. Summary diagnostic performance metrics and their 95% confidence intervals (CI) were calculated.

RESULTS: The study included 109 articles permitting analyses for lung (n = 34), liver (n = 29), colorectal (n = 28), pancreatic (n = 16), breast (n = 17), esophageal (n = 12), ovarian (n = 13), gastric (n = 9), head and neck (n = 4), and mixed (n = 27) cancer types. Specificity was consistently high across all tumor types and stages (94%-99%). Sensitivity ranged from 72% to 92% for stage I-III, 44-91% for stage I, 71-98% for stage II and 83-99% for stage III. In the pooled study population, neural networks (90%, 95% CI: 81%-95%), random forest (86%, 95% CI: 77%-92%) and heterogeneous ensemble learning (85%, 95% CI: 79%-89%) demonstrated the highest sensitivity. The stratified analysis by classifier feature revealed 86% (95% CI: 80%-90%) sensitivity for fragmentation and 81% (95% CI: 76%-85%) for methylation, with 92%-96% specificity.

CONCLUSION: ML and cfDNA profiling show potential for early cancer detection, with ensemble methods, neural networks and random forests achieving the best overall performance. Fragmentomic features provide the highest sensitivity.

PMID:41930854 | DOI:10.1016/j.ejca.2026.116699

  •  
  • ↗️
↗️

Integrating liquid biopsies and artificial intelligence for early cancer detection: A systematic review and meta-analysis

✇(Multiomics OR Omics) AND (Lung OR gastric OR Hepatocellular)
By: Panagiotis Filis · Georgios Markozannes · Dimitrios Salgkamis · Nikos Tsiknakis · Ioannis Zerdes · Eirini Pagkalidou · Maria Manou · Nikolaos Filis · Andri Papakonstantinou · Dimitrios Mavroudis · Alexios Matikas · Konstantinos K Tsilidis · Theodoros Foukakis

Eur J Cancer. 2026 Mar 24;239:116699. doi: 10.1016/j.ejca.2026.116699. Online ahead of print.

ABSTRACT

INTRODUCTION: The latest generation of liquid biopsies incorporates multi-omic features, including genomics, methylomics, and fragmentomics. Machine learning (ML) approaches have been proposed to synthesize these complex biological data for the development of diagnostic classifiers. This study aims to evaluate the integration of ML with circulating cell-free DNA (cfDNA) analysis for early cancer detection.

METHODS: Medline, Embase, Cochrane, and Web of Science were searched in July 2025. Eligible studies combined ML and cfDNA features to distinguish cancer patients (stages I-III) from non-cancer controls. Summary diagnostic performance metrics and their 95% confidence intervals (CI) were calculated.

RESULTS: The study included 109 articles permitting analyses for lung (n = 34), liver (n = 29), colorectal (n = 28), pancreatic (n = 16), breast (n = 17), esophageal (n = 12), ovarian (n = 13), gastric (n = 9), head and neck (n = 4), and mixed (n = 27) cancer types. Specificity was consistently high across all tumor types and stages (94%-99%). Sensitivity ranged from 72% to 92% for stage I-III, 44-91% for stage I, 71-98% for stage II and 83-99% for stage III. In the pooled study population, neural networks (90%, 95% CI: 81%-95%), random forest (86%, 95% CI: 77%-92%) and heterogeneous ensemble learning (85%, 95% CI: 79%-89%) demonstrated the highest sensitivity. The stratified analysis by classifier feature revealed 86% (95% CI: 80%-90%) sensitivity for fragmentation and 81% (95% CI: 76%-85%) for methylation, with 92%-96% specificity.

CONCLUSION: ML and cfDNA profiling show potential for early cancer detection, with ensemble methods, neural networks and random forests achieving the best overall performance. Fragmentomic features provide the highest sensitivity.

PMID:41930854 | DOI:10.1016/j.ejca.2026.116699

  •  
  • ↗️
↗️

Integrating liquid biopsies and artificial intelligence for early cancer detection: A systematic review and meta-analysis

✇Omics in Gastric
By: Panagiotis Filis · Georgios Markozannes · Dimitrios Salgkamis · Nikos Tsiknakis · Ioannis Zerdes · Eirini Pagkalidou · Maria Manou · Nikolaos Filis · Andri Papakonstantinou · Dimitrios Mavroudis · Alexios Matikas · Konstantinos K Tsilidis · Theodoros Foukakis

Eur J Cancer. 2026 Mar 24;239:116699. doi: 10.1016/j.ejca.2026.116699. Online ahead of print.

ABSTRACT

INTRODUCTION: The latest generation of liquid biopsies incorporates multi-omic features, including genomics, methylomics, and fragmentomics. Machine learning (ML) approaches have been proposed to synthesize these complex biological data for the development of diagnostic classifiers. This study aims to evaluate the integration of ML with circulating cell-free DNA (cfDNA) analysis for early cancer detection.

METHODS: Medline, Embase, Cochrane, and Web of Science were searched in July 2025. Eligible studies combined ML and cfDNA features to distinguish cancer patients (stages I-III) from non-cancer controls. Summary diagnostic performance metrics and their 95% confidence intervals (CI) were calculated.

RESULTS: The study included 109 articles permitting analyses for lung (n = 34), liver (n = 29), colorectal (n = 28), pancreatic (n = 16), breast (n = 17), esophageal (n = 12), ovarian (n = 13), gastric (n = 9), head and neck (n = 4), and mixed (n = 27) cancer types. Specificity was consistently high across all tumor types and stages (94%-99%). Sensitivity ranged from 72% to 92% for stage I-III, 44-91% for stage I, 71-98% for stage II and 83-99% for stage III. In the pooled study population, neural networks (90%, 95% CI: 81%-95%), random forest (86%, 95% CI: 77%-92%) and heterogeneous ensemble learning (85%, 95% CI: 79%-89%) demonstrated the highest sensitivity. The stratified analysis by classifier feature revealed 86% (95% CI: 80%-90%) sensitivity for fragmentation and 81% (95% CI: 76%-85%) for methylation, with 92%-96% specificity.

CONCLUSION: ML and cfDNA profiling show potential for early cancer detection, with ensemble methods, neural networks and random forests achieving the best overall performance. Fragmentomic features provide the highest sensitivity.

PMID:41930854 | DOI:10.1016/j.ejca.2026.116699

  •  
  • ↗️
↗️

Integrating liquid biopsies and artificial intelligence for early cancer detection: A systematic review and meta-analysis

✇Omics In Lung
By: Panagiotis Filis · Georgios Markozannes · Dimitrios Salgkamis · Nikos Tsiknakis · Ioannis Zerdes · Eirini Pagkalidou · Maria Manou · Nikolaos Filis · Andri Papakonstantinou · Dimitrios Mavroudis · Alexios Matikas · Konstantinos K Tsilidis · Theodoros Foukakis

Eur J Cancer. 2026 Mar 24;239:116699. doi: 10.1016/j.ejca.2026.116699. Online ahead of print.

ABSTRACT

INTRODUCTION: The latest generation of liquid biopsies incorporates multi-omic features, including genomics, methylomics, and fragmentomics. Machine learning (ML) approaches have been proposed to synthesize these complex biological data for the development of diagnostic classifiers. This study aims to evaluate the integration of ML with circulating cell-free DNA (cfDNA) analysis for early cancer detection.

METHODS: Medline, Embase, Cochrane, and Web of Science were searched in July 2025. Eligible studies combined ML and cfDNA features to distinguish cancer patients (stages I-III) from non-cancer controls. Summary diagnostic performance metrics and their 95% confidence intervals (CI) were calculated.

RESULTS: The study included 109 articles permitting analyses for lung (n = 34), liver (n = 29), colorectal (n = 28), pancreatic (n = 16), breast (n = 17), esophageal (n = 12), ovarian (n = 13), gastric (n = 9), head and neck (n = 4), and mixed (n = 27) cancer types. Specificity was consistently high across all tumor types and stages (94%-99%). Sensitivity ranged from 72% to 92% for stage I-III, 44-91% for stage I, 71-98% for stage II and 83-99% for stage III. In the pooled study population, neural networks (90%, 95% CI: 81%-95%), random forest (86%, 95% CI: 77%-92%) and heterogeneous ensemble learning (85%, 95% CI: 79%-89%) demonstrated the highest sensitivity. The stratified analysis by classifier feature revealed 86% (95% CI: 80%-90%) sensitivity for fragmentation and 81% (95% CI: 76%-85%) for methylation, with 92%-96% specificity.

CONCLUSION: ML and cfDNA profiling show potential for early cancer detection, with ensemble methods, neural networks and random forests achieving the best overall performance. Fragmentomic features provide the highest sensitivity.

PMID:41930854 | DOI:10.1016/j.ejca.2026.116699

  •  
  • ↗️
↗️

Analysis of LLM Performance on AWS Bedrock: Receipt-item Categorisation Case Study

✇cs.AI, q-bio.NC updates on arXiv.org
By: Gabby Sanchez · Sneha Oommen · Cassandra T. Britto · Di Wang · Jung-De Chiou · Maria Spichkova
arXiv:2604.01615v1 Announce Type: new Abstract: This paper presents a systematic, cost-aware evaluation of large language models (LLMs) for receipt-item categorisation within a production-oriented classification framework. We compare four instruction-tuned models available through AWS Bedrock: Claude 3.7 Sonnet, Claude 4 Sonnet, Mixtral 8x7B Instruct, and Mistral 7B Instruct. The aim of the study was (1) to assess performance across accuracy, response stability, and token-level cost, and (2) to investigate what prompting methods, zero-shot or few-shot, are especially appropriate both in terms of accuracy and in terms of incurred costs. Results of our experiments demonstrated that Claude 3.7 Sonnet achieves the most favourable balance between classification accuracy and cost efficiency.
  •  
  • ↗️
↗️

Retrieval-aligned Tabular Foundation Models Enable Robust Clinical Risk Prediction in Electronic Health Records Under Real-world Constraints

✇cs.AI, q-bio.NC updates on arXiv.org
By: Minh-Khoi Pham · Thang-Long Nguyen Ho · Thao Thi Phuong Dao · Tai Tan Mai · Minh-Triet Tran · Marie E. Ward · Una Geary · Rob Brennan · Nick McDonald · Martin Crane · Marija Bezbradica
arXiv:2604.01841v1 Announce Type: new Abstract: Clinical prediction from structured electronic health records (EHRs) is challenging due to high dimensionality, heterogeneity, class imbalance, and distribution shift. While tabular in-context learning (TICL) and retrieval-augmented methods perform well on generic benchmarks, their behavior in clinical settings remains unclear. We present a multi-cohort EHR benchmark comparing classical, deep tabular, and TICL models across varying data scale, feature dimensionality, outcome rarity, and cross-cohort generalization. PFN-based TICL models are sample-efficient in low-data regimes but degrade under naive distance-based retrieval as heterogeneity and imbalance increase. We propose AWARE, a task-aligned retrieval framework using supervised embedding learning and lightweight adapters. AWARE improves AUPRC by up to 12.2% under extreme imbalance, with gains increasing with data complexity. Our results identify retrieval quality and retrieval-inference alignment as key bottlenecks for deploying tabular in-context learning in clinical prediction.
  •  
  • ↗️
↗️

Bayesian Elicitation with LLMs: Model Size Helps, Extra "Reasoning" Doesn't Always

✇cs.AI, q-bio.NC updates on arXiv.org
By: Luka Hobor · Mario Brcic · Mihael Kovac · Kristijan Poje
arXiv:2604.01896v1 Announce Type: new Abstract: Large language models (LLMs) have been proposed as alternatives to human experts for estimating unknown quantities with associated uncertainty, a process known as Bayesian elicitation. We test this by asking eleven LLMs to estimate population statistics, such as health prevalence rates, personality trait distributions, and labor market figures, and to express their uncertainty as 95\% credible intervals. We vary each model's reasoning effort (low, medium, high) to test whether more "thinking" improves results. Our findings reveal three key results. First, larger, more capable models produce more accurate estimates, but increasing reasoning effort provides no consistent benefit. Second, all models are severely overconfident: their 95\% intervals contain the true value only 9--44\% of the time, far below the expected 95\%. Third, a statistical recalibration technique called conformal prediction can correct this overconfidence, expanding the intervals to achieve the intended coverage. In a preliminary experiment, giving models web search access degraded predictions for already-accurate models, while modestly improving predictions for weaker ones. Models performed well on commonly discussed topics but struggled with specialized health data. These results indicate that LLM uncertainty estimates require statistical correction before they can be used in decision-making.
  •  
  • ↗️
↗️

GenGait: A Transformer-Based Model for Human Gait Anomaly Detection and Normative Twin Generation

✇cs.AI, q-bio.NC updates on arXiv.org
By: Elisa Motta · Marta Lorenzini · Clara Mouawad · Alberto Ranavolo · Mariano Serrao · Arash Ajoudani
arXiv:2604.01997v1 Announce Type: new Abstract: Gait analysis provides an objective characterization of locomotor function and is widely used to support diagnosis and rehabilitation monitoring across neurological and orthopedic disorders. Deep learning has been increasingly applied to this domain, yet most approaches rely on supervised classifiers trained on disease-labeled data, limiting generalization to heterogeneous pathological presentations. This work proposes a label-free framework for joint-level anomaly detection and kinematic correction based on a Transformer masked autoencoder trained exclusively on normative gait sequences from 150 adults, acquired with a markerless multi-camera motion-capture system. At inference, a two-pass procedure is applied to potentially pathological input sequences, first it estimates joint inconsistency scores by occluding individual joints and measuring deviations from the learned normative prior. Then, it withholds the flagged joints from the encoder input and reconstructs the full skeleton from the remaining spatiotemporal context, yielding corrected kinematic trajectories at the flagged positions. Validation on 10 held-out normative participants, who mimicked seven simulated gait abnormalities, showed accurate localization of biomechanically inconsistent joints, a significant reduction in angular deviation across all analyzed joints with large effect sizes, and preservation of normative kinematics. The proposed approach enables interpretable, subject-specific localization of gait impairments without requiring disease labels. Video is available at https://youtu.be/Rcm3jqR5pN4.
  •  
  • ↗️
↗️

Phase estimation with autoregressive padding (PEAP): addressing inaccuracies and biases in EEG analysis

✇cs.AI, q-bio.NC updates on arXiv.org
By: Miriam Kirchhoff, Johanna R\"osch, Maria Ermolova, Oskari Ahola, Sarah Harders, Juliana Hougland, Ulf Ziemann
arXiv:2604.02212v1 Announce Type: new Abstract: Accurate phase estimation at the edge of data segments is crucial for EEG applications such as EEG-TMS in offline and real-time data analysis. Our research evaluates the phase estimation performance of four commonly used methods (Phastimate, SSPE, ETP, and PhastPadding) for accuracy and systemic biases, using data from young and elderly healthy controls and chronic stroke participants. To address the identified limitations of the established methods, we introduce Phase Estimation with Autoregressive Padding (PEAP), a method that prevents strong bandpass filtering-induced artifacts. Contrary to the established methods, PEAP does not show significant biases and improves accuracy by 3.2 to 9.2% for the continuous phase estimation. Our offline analysis demonstrates how established methods are systematically biased towards some estimates and how they induce phase shifts. We also show that differences between methods do not vary between clinical and control populations, supporting their translatability. This work indicates that systematic biases in established phase estimation methods may compromise the validity and comparability of phase-dependent findings. PEAP addresses these limitations and thus offers a more reliable and more accurate alternative method.
  •  
  • ↗️
↗️

Transformer self-attention encoder-decoder with multimodal deep learning for response time series forecasting and digital twin support in wind structural health monitoring

✇cs.AI, q-bio.NC updates on arXiv.org
By: Feiyu Zhou · Marios Impraimakis
arXiv:2604.01712v1 Announce Type: cross Abstract: The wind-induced structural response forecasting capabilities of a novel transformer methodology are examined here. The model also provides a digital twin component for bridge structural health monitoring. Firstly, the approach uses the temporal characteristics of the system to train a forecasting model. Secondly, the vibration predictions are compared to the measured ones to detect large deviations. Finally, the identified cases are used as an early-warning indicator of structural change. The artificial intelligence-based model outperforms approaches for response forecasting as no assumption on wind stationarity or on structural normal vibration behavior is needed. Specifically, wind-excited dynamic behavior suffers from uncertainty related to obtaining poor predictions when the environmental or traffic conditions change. This results in a hard distinction of what constitutes normal vibration behavior. To this end, a framework is rigorously examined on real-world measurements from the Hardanger Bridge monitored by the Norwegian University of Science and Technology. The approach captures accurate structural behavior in realistic conditions, and with respect to the changes in the system excitation. The results, importantly, highlight the potential of transformer-based digital twin components to serve as next-generation tools for resilient infrastructure management, continuous learning, and adaptive monitoring over the system's lifecycle with respect to temporal characteristics.
  •  
  • ↗️
↗️

A deep learning pipeline for PAM50 subtype classification using histopathology images and multi-objective patch selection

✇cs.AI, q-bio.NC updates on arXiv.org
By: Arezoo Borji · Gernot Kronreif · Bernhard Angermayr · Francisco Mario Calisto · Wolfgang Birkfellner · Inna Servetnyk · Yinyin Yuan · Sepideh Hatamikia
arXiv:2604.01798v1 Announce Type: cross Abstract: Breast cancer is a highly heterogeneous disease with diverse molecular profiles. The PAM50 gene signature is widely recognized as a standard for classifying breast cancer into intrinsic subtypes, enabling more personalized treatment strategies. In this study, we introduce a novel optimization-driven deep learning framework that aims to reduce reliance on costly molecular assays by directly predicting PAM50 subtypes from H&E-stained whole-slide images (WSIs). Our method jointly optimizes patch informativeness, spatial diversity, uncertainty, and patch count by combining the non-dominated sorting genetic algorithm II (NSGA-II) with Monte Carlo dropout-based uncertainty estimation. The proposed method can identify a small but highly informative patch subset for classification. We used a ResNet18 backbone for feature extraction and a custom CNN head for classification. For evaluation, we used the internal TCGA-BRCA dataset as the training cohort and the external CPTAC-BRCA dataset as the test cohort. On the internal dataset, an F1-score of 0.8812 and an AUC of 0.9841 using 627 WSIs from the TCGA-BRCA cohort were achieved. The performance of the proposed approach on the external validation dataset showed an F1-score of 0.7952 and an AUC of 0.9512. These findings indicate that the proposed optimization-guided, uncertainty-aware patch selection can achieve high performance and improve the computational efficiency of histopathology-based PAM50 classification compared to existing methods, suggesting a scalable imaging-based replacement that has the potential to support clinical decision-making.
  •  
  • ↗️
↗️

Tracking the emergence of linguistic structure in self-supervised models learning from speech

✇cs.AI, q-bio.NC updates on arXiv.org
By: Marianne de Heer Kloots · Martijn Bentum · Hosein Mohebbi · Charlotte Pouw · Gaofei Shen · Willem Zuidema
arXiv:2604.02043v1 Announce Type: cross Abstract: Self-supervised speech models learn effective representations of spoken language, which have been shown to reflect various aspects of linguistic structure. But when does such structure emerge in model training? We study the encoding of a wide range of linguistic structures, across layers and intermediate checkpoints of six Wav2Vec2 and HuBERT models trained on spoken Dutch. We find that different levels of linguistic structure show notably distinct layerwise patterns as well as learning trajectories, which can partially be explained by differences in their degree of abstraction from the acoustic signal and the timescale at which information from the input is integrated. Moreover, we find that the level at which pre-training objectives are defined strongly affects both the layerwise organization and the learning trajectories of linguistic structures, with greater parallelism induced by higher-order prediction tasks (i.e. iteratively refined pseudo-labels).
  •  
  • ↗️
↗️

Four Generations of Quantum Biomedical Sensors

✇cs.AI, q-bio.NC updates on arXiv.org
By: Xin Jin · Priyam Srivastava · Ronghe Wang · Yuqing Li · Jonathan Beaumariage · Tom Purdy · M. V. Gurudev Dutt · Kang Kim · Kaushik Seshadreesan · Junyu Liu
arXiv:2603.29944v2 Announce Type: replace-cross Abstract: Quantum sensing technologies offer transformative potential for ultra-sensitive biomedical sensing, yet their clinical translation remains constrained by classical noise limits and a reliance on macroscopic ensembles. We propose a unifying generational framework to organize the evolving landscape of quantum biosensors based on their utilization of quantum resources. First-generation devices utilize discrete energy levels for signal transduction but follow classical scaling laws. Second-generation sensors exploit quantum coherence to reach the standard quantum limit, while third-generation architectures leverage entanglement and spin squeezing to approach Heisenberg-limited precision. We further define an emerging fourth generation characterized by the end-to-end integration of quantum sensing with quantum learning and variational circuits, enabling adaptive inference directly within the quantum domain. By analyzing critical parameters such as bandwidth matching and sensor-tissue proximity, we identify key technological bottlenecks and propose a roadmap for transitioning from measuring physical observables to extracting structured biological information with quantum-enhanced intelligence.
  •  
  • ↗️
↗️

The exposome of brain aging across 34 countries

✇Nature Medicine
By: Agustina Legaz · Sebastian Moguilner · Pablo Barttfeld · Jhosmary Cuadros Castro · Dante Sebastián Galván Rial · Joaquin Migeot · Francesca R. Farina · Raul Gonzalez-Gomez · Salomón Salazar-Londoño · Tiago de Souza Azzi · Sarael Alcauter · Lucia Amoruso · Renato Anghinah · May Bakr · María Isabel Behrens · Martin A. Bruno · Juan Felipe Cardona · Christopher Chen · Olivier Colliot · Nilton Custodio · Quentin d’Acremont · Stéphanie Debette · Patrycja Dzianok · Bahar Güntekin · Jordi Huguet · Cecilia Jarne · Marc Joliot · Elissaios Karageorgiou · Eman M. Khedr · Sheri-Michelle Koopowitz · Ewa Kublik · Jinkook Lee · Leon Aksman · Susanna Lopez · Paula Margaretic · Diana Matallana · Sophie Matis · Andreas Miltiadous · Gabriela Novotni · Patricio Orio · Mario A. Parra · Simone Reppermund · Elisa de Paula França Resende · Giovanni Abrahão Salum · Andrea Slachevsky · Ana Luisa Sosa Ortiz · Leonel Tadao Takada · An Qi Toh · Alexandros Tzallas · Wei Wen · Robert Whelan · Görsev Yener · Jonathan Adrián Zegarra Valdivia · Athanasia Alexoudi · Faheem Arshad · Ha

Nature Medicine, Published online: 03 April 2026; doi:10.1038/s41591-026-04302-z

Exposome analyses across 34 countries showed that social exposures were associated with faster functional brain aging and physical exposures with faster structural brain aging.
  •  
  • ↗️
↗️

Artificial Intelligence, Connected Care, and Enabling Digital Health Technologies in Rare Diseases With a Focus on Lysosomal Storage Disorders: Scoping Review

✇Journal of Medical Internet Research
By: Alberta MC Spreafico · Luca Neri · Kim Angel · Cinzia Maria Bellettato · Maurizio Scarpa
Background: Rare diseases affect more than 300 million people globally, and only about 5% have approved therapies. Lysosomal storage disorders (LSDs) exemplify the diagnostic and long-term care complexity typical of rare diseases, and digital health technologies (DHTs), especially artificial intelligence (AI) and connected care (CC), are emerging tools to support LSD management. Objective: We aimed to map and synthesize peer-reviewed and gray literature from the past decade on DHTs relevant for LSD care, with a primary analytic focus on AI-enabled and CC solutions and a contextual mapping of other enabling DHTs. Evidence distribution was charted by population, care-journey phase, and outcome domains to identify gaps, methodological limitations, and timely priorities relevant for research, clinical practice implementation, and policies. Methods: We conducted a scoping review guided by a population, concept, context framework and operationalized through a Population, Intervention, Comparison, and Outcome (PICO)-informed data-charting structure to map study characteristics and reported outcomes, without causal or effectiveness assumptions and without risk-of-bias assessment. We searched PubMed, Google Scholar, and ClinicalTrials.gov for studies published between October 2015 and September 2024, complemented by AI-assisted discovery tools for citation extension. Reproducibility logs (search strings, run dates, filters, and stepwise counts) were maintained. Of 1751 records retrieved, 245 were included. Evidence was charted by LSD population, intervention class (AI, CC, and other enabling DHTs), outcome domains (patient, health care, and societal), and phase of the care journey. Results: Among 245 included records, 92.2% (226/245) were peer-reviewed, and 7.8% (19/245) were gray literature; no completed and published randomized controlled trials or LSD-specific systematic reviews were identified, with evidence dominated by small, single-center observational studies. Overall, 40 peer-reviewed records reported AI-driven DHTs, 89 reported CC DHTs, and 144 reported other enabling DHTs (some multilabeled). Evidence was concentrated mostly in Gaucher and Fabry diseases. Nearly half of the mapped literature focused on screening and diagnosis, with fewer records addressing treatment intensification, rehabilitation, and end-of-life care. Outcomes were predominantly health care delivery performance measures, with fewer patient and societal outcomes. AI applications mainly supported diagnostic decision support, phenotyping, monitoring, tracking, and risk stratification; CC commonly involved telemedicine, remote monitoring, and patient-engagement platforms; enabling DHTs included interoperable data systems, registries, and digital infrastructures. Conclusions: The evidence base is appreciable for a niche field and reflects growing interest in AI and CC for LSD care, but heterogeneity and methodological limitations preclude inferences on effectiveness or routine implementation. This evidence map highlights relatively stronger areas and gaps, providing a structured foundation to inform timely expert consensus-building and research prioritization. Key priorities include interoperable data infrastructures and data availability, prospective multicenter evaluations, transparent reporting of algorithms and workflows, and implementation-relevant outcomes to support safe, equitable, and scalable adoption aligned with evolving European Union and global rare-disease priorities.
  •  
  • ↗️
↗️

Effectiveness of the Components of a Digital Multiple Health Behavior Change Intervention Among Individuals Seeking Help Online (Coach): Factorial Randomized Trial

✇Journal of Medical Internet Research
By: Joel Crawford · Katarina Åsberg · Jenny Blomqvist · Oskar Lundgren · Hanna Henriksson · Pontus Henriksson · Preben Bendtsen · Marie Löf · Marcus Bendtsen
Background: Extant digital multiple health behavior change interventions have shown promise in various populations; however, evidence for a broader approach among the general population is lacking. Moreover, existing interventions often contain several components but are typically assessed as a whole, meaning it remains unclear to what extent individual components contribute to intervention effects and how they may interact to influence health outcomes. Objective: This study estimates the effects of 6 components of a digital health behavior change intervention on alcohol, diet, physical activity, and smoking outcomes among individuals searching for help online. Methods: A double-blind randomized factorial trial design with 6 two-level factors was used. Adults from the general public in Sweden who were seeking help to change their behaviors were recruited through web searches and social media. Participants were eligible if they were 18 years or older and had at least one health behavior classified as unhealthy. Effects of 6 components were estimated: screening/feedback, goal-setting/planning, motivation, skills/know-how, mindfulness, and self-authored SMS text messages. Primary outcomes were weekly alcohol consumption and frequency of heavy episodic drinking, average daily fruit and vegetable consumption, weekly moderate-to-vigorous physical activity, and 4-week point-prevalence smoking. Results: A total of 5419 individuals were randomized. Overall, the screening/feedback component was the most effective for changing health behaviors, along with goal-setting/planning and motivation to change. In particular, there was evidence that screening/feedback increased average daily portions of fruit and vegetables at 2 months (mean difference 0.17, compatibility interval [CoI] 0.09-0.25, probability of effect [POE] >99.9%) and at 4 months (mean difference 0.13, CoI 0.04-0.21, POE 99.9%) and reduced the frequency of heavy episodic drinking at 4 months (incidence rate ratio 0.91, CoI 0.81-1.03, POE 94.2%). Components also interacted to further improve health outcomes, most notably the combination of screening/feedback with motivation to change, which further increased fruit and vegetable consumption (2 months: mean difference 0.20, CoI 0.09-0.30, POE >99.9%; 4 months: mean difference 0.17, CoI 0.05-0.29, POE 99.8%). Conclusions: The results from this study contribute to the development of more effective interventions by providing novel insights into the effects of individual and pairwise components of complex digital health behavior change interventions. Trial Registration: ISRCTN Registry ISRCTN16420548; http://www.isrctn.com/ISRCTN16420548
  •  
  • ↗️
↗️

Artificial Intelligence in Health Professions Education: Qualitative Study of Student Experiences

✇Journal of Medical Internet Research
By: Marina Guirguis · Salomon Fotsing · Joanne Fevry · Christine Landry · Diane Bouchard-Lamothe · Jennifer Lacroix · Alireza Jalali
Background: Artificial intelligence (AI) is increasingly integrated into education and health care, raising questions about how students use these technologies and how AI influences their learning. In health education, understanding these trends is particularly important because student learning directly impacts future clinical skills. Objective: This study aimed to explore the use of AI tools by health sciences students at the University of Ottawa. More specifically, it sought to identify the most frequently used AI tools, describe students’ usage habits, determine which tools support knowledge acquisition and skill development, and gather students’ recommendations for effective strategies to raise awareness and train their peers on the responsible use of AI. Methods: A qualitative approach was used with students from 10 health professions who reported using AI in their studies. Data were collected through semistructured interviews and an open-ended qualitative online survey. Inductive thematic analysis within an interpretive paradigm was applied to capture patterns, perceptions, and emergent themes. Results: A total of 51 health professions students participated in the study. Most were women between the ages of 20 and 29 years. ChatGPT (OpenAI) emerged as the most frequently used AI tool. Students perceived AI as a complementary tool that facilitated knowledge acquisition, skill development, writing, and problem-solving. AI adoption was driven by curiosity, peer influence, and the desire to improve work efficiency. Students critically evaluated AI results, integrated the tools into their learning processes, and emphasized the importance of technical skills, critical thinking, and digital literacy. Peer learning, hands-on demonstrations, and access to online resources were recommended for effective AI training. Conclusions: This research demonstrates that health professions students actively use AI tools, particularly ChatGPT, to support learning, skill development, and academic tasks. Although AI is valuable as an educational aid and its use varies by student and context, this highlights the need for structured guidance, critical evaluation skills, and peer-supported training. These findings highlight the importance of thoughtfully integrating AI into educational programs to enhance learning outcomes, foster skill acquisition, and ensure responsible and effective adoption.
  •  
  • ↗️
↗️

Semaglutide on liver fibrosis and heart outcomes in patients at high risk of liver fibrosis: a prespecified analysis of the SELECT randomized trial

✇Nature Medicine
By: Sebastian M. Meyhöfer · Bertrand Cariou · Cintia Cercato · Helen M. Colhoun · John Deanfield · Michelle T. Long · Ole Kleist Jeppesen · A. Michael Lincoff · Ildiko Lingvay · Jorge Plutzky · Philip N. Newsome · Stephen J. Nicholls · Maria Quiroga · Ferruccio Santini · Arun J. Sanyal · Steven E. Kahn · Donna H. Ryan · Scott S. Emerson · Robert F. Kushner · Kirstine Brown-Frandsen · G. Kees Hovingh · Søren Hardt-Lindberg · Christoffer W. Tornøe

Nature Medicine, Published online: 02 April 2026; doi:10.1038/s41591-026-04281-1

A prespecified analysis from the SELECT trial showed that semaglutide reduces major adverse cardiovascular events by 20% compared with placebo, particularly in patients at high risk of fibrosis, as indicated by the Fibrosis-4 index.
  •  
  • ↗️
❌