Journal of Medical Systems最新文献

Chronic disease management requires continuous monitoring, lifestyle modification and therapy adherence, thus requiring constant support from healthcare professionals. Chatbots have proven to be a promising approach for engaging patients in managing their health condition at home and for offering continuous assistance by being readily available to answer questions. While large language models offer an impressive solution for chatbot implementation, third-party systems raise privacy concerns, and computational requirements limit small-scale deployment. We address these challenges by developing a chatbot for hypertensive patients based on open-source small language models (SLMs), specifically designed for running on personal resource-constrained devices and for providing assistance in QA tasks. In order to guarantee comparable conversational performances with respect to larger language models, we exploited retrieval-augmented generation (RAG) with a local knowledge base. This ensures data privacy by deploying models locally while achieving competitive accuracy and maintaining low computational costs suitable for end-user devices. We experimented with eight SLMs, two prompt configurations, and different RAG strategies - both in the embedding and retrieval components - to identify the most effective solution. The evaluation of our solution grounds on both reference metrics and expert evaluation. Our findings suggest that RAG-enhanced SLMs can improve response clarity and content accuracy. However, our results also indicate that newer SLMs like Qwen3 demonstrate strong performance even without RAG, suggesting a potential shift in the necessity for complex retrieval mechanisms with rapidly evolving model architectures.

The cost-effective open-source artificial intelligence (AI) model DeepSeek-R1 in China holds significant potential for healthcare applications. As a health education tool, it could help patients acquire health science knowledge and improve health literacy. Low back pain (LBP), the most common musculoskeletal problem globally, has seen increasing use of large language model (LLM)-based AI chatbots by patients to access health information, making it critical to further examine the quality of such information. This study aimed to evaluate the response quality and readability of answers generated by DeepSeek-R1 to common patient questions about LBP. Ten questions were formulated using inductive methods based on literature analysis and Baidu Index data, which were presented to DeepSeek-R1 on March 10, 2025. The evaluation spanned readability, understandability, actionability, clinician assessment, and reference assessment. Readability was measured using the Flesch-Kincaid Grade Level, Flesch Reading Ease Scale, Gunning Fog Index, Coleman-Liau Index, and Simple Measure of Gobbledygook (SMOG Index). Understandability and actionability were assessed via the Patient Education Materials and Assessment Tool for Printable Materials (PEMAT-P). Clinicians evaluated accuracy, completeness, and correlation. A reference evaluation tool was used to assess reference quality and the presence of hallucinations. Readability analysis indicated that DeepSeek's responses were overall "difficult to read", with Flesch-Kincaid Grade Level (mean 12.39, SD 1.91), Flesch Reading Ease Scale (mean 19.55, Q1 12.94, Q3 29.78), Gunning Fog Index (mean 13.95, SD 2.61), Coleman-Liau Index (mean 17.46, SD 2.30), and SMOG Index (mean 11.04, SD 1.37). PEMAT-P revealed good understandability but weak actionability. Consensus among five clinicians confirmed satisfactory accuracy, completeness, and relevance. References Assessment identified 9 instances (14.8%) of hallucinated references, while Supporting was rated as moderate, with most references sourced from authoritative platforms. Our study demonstrates the potential of DeepSeek-R1 in the educational content for patients with LBP. It can be employed as a supplement to patient education tools rather than substituting for clinical judgment.

Zero-shot anomaly detection (ZSAD) is gaining traction in medical imaging as a way to identify abnormalities without task-specific supervision. In this work, we benchmark state-of-the-art CLIP-based ZSAD models -originally developed for industrial inspection -on brain metastasis detection using the BraTS-MET dataset. We evaluate both general-purpose and medical-adapted variants across multiple training paradigms with little to no supervision, emulating real-world scenarios with scarce labeled imaging data. While the models can apply general knowledge to medical images, we show that their accuracy remains limited, especially in peripheral brain regions, and that substantial but still suboptimal performance gains are achieved only via domain-specific fine-tuning. Our findings highlight current limitations in spatial consistency when using 2D-based approaches for 3D problems, and suggest that adaptation is required to make CLIP-based ZSAD viable for clinical use.

Background: Assessing Left Ventricular Ejection Fraction (LVEF) is crucial for diagnosing reduced systolic function, yet echocardiography (ECHO) may not always be readily available, potentially delaying treatment. Electrocardiography (ECG) offers a cost-effective and accessible alternative for estimating LVEF. However, specialized AI models for this purpose are often complex and costly to develop.

Objective: This study uniquely evaluates GPT-4o Fine-Tuned Vision (GPT-4o-FTV), a general-purpose AI model, for detecting LVEF ≤ 35% from ECG images, comparing its performance with a Convolutional Neural Network (CNN) model and clinician assessments.

Methods: We analyzed ECGs from 202 patients (42.6% women, mean age 64.5 ± 16.3 years) at a tertiary center, excluding those with pacemakers and including only high-quality ECGs. LVEF ≤ 35% was present in 11.9% (n = 24). GPT-4o-FTV, trained on 20 labeled ECGs, was tested using a structured prompt across four runs. Accuracy, sensitivity, specificity, and positive predictive value (PPV) were compared to a CNN model and four clinicians.

Results: GPT-4o-FTV achieved 79.9% accuracy, 72.9% sensitivity, 80.8% specificity, an F1-score of 46.4%, and a PPV of 34%, outperforming clinicians (74.9% accuracy, 65.6% sensitivity, 76.1% specificity, 39% F1-score, PPV 27.9%). The CNN model had the highest performance (89.1% accuracy, 79.2% sensitivity, 90.4% specificity, 63.3% F1-score, PPV 52.8%).

Conclusions: GPT-4o-FTV demonstrates strong potential as an accessible tool for cardiac diagnostics, particularly in resource-limited settings. While CNN models remain superior in accuracy, the ease of fine-tuning GPT-4o-FTV highlights its practical utility. Future research should focus on larger datasets, additional optimization, and exploring its ability to detect early predictors of LVEF decline.

Public health increasingly relies on digital infrastructures, yet data remains fragmented across clinical, behavioral, and social domains. Customer Data Platforms (CDPs), originally created in marketing to unify diverse information into dynamic individual profiles, could provide a new approach for person-centered public health. This article explores the strategic potential of applying CDP principles, such as data unification, identity resolution, segmentation, and timely intervention, to enhance surveillance, prevention, and chronic disease management. A conceptual framework is presented and demonstrated through a breast cancer screening scenario, illustrating how CDPs could enable personalized outreach and integration with artificial intelligence (AI). Although promising, there are significant challenges related to privacy, interoperability, fairness, and governance. Responsible deployment requires socio-technical strategies that emphasize transparency, ethical oversight, and person involvement.

Electrocorticography (ECoG) signals provide a valuable window into neural activity, yet their complex structure makes reliable classification challenging. This study addresses the problem by proposing a feature-selective framework that integrates multiple feature extraction techniques with statistical feature selection to improve classification performance. Power spectral density, wavelet-based features, Shannon entropy, and Hjorth parameters were extracted from ECoG signals obtained during a visual task. The most informative features were then selected using analysis of variance (ANOVA), and classification was performed with several machine learning methods, including decision trees, support vector machines, neural networks, and long short-term memory (LSTM) networks. Experimental results show that the proposed framework achieves high accuracy across individual patients as well as the combined dataset, with clear separability between classes confirmed through t-SNE visualization. In addition, analysis of selected features highlights the prominent role of electrodes located near the visual cortex, providing insights into the spatial distribution of neural activity.

Motor imagery (MI) is widely used in brain-computer interfaces (BCIs) due to its simplicity and reproducibility, enabling individuals with motor impairments to perform non-muscular limb training for the rehabilitation of motor-related neurons. While MI-based BCIs have shown promise for neurorehabilitation, current 2D paradigms fail to engage critical sensorimotor networks. To address this limitation, we designed an immersive MI paradigm in a virtual reality (VR) environment, where participants imagined limb movements in response to continuous three-dimensional (3D) palm motion stimuli. Furthermore, we proposed a novel decoding algorithm that integrates depthwise separable convolution with multi-head self-attention mechanisms. The proposed method was evaluated against existing approaches, demonstrating superior classification accuracy while reducing the temporal and spatial complexity associated with attention mechanisms. To assess the generalizability and robustness of the algorithm across different scenarios, we conducted experiments on two publicly available datasets: BCI Competition IV-2a and the PhysioNet MI dataset. Results showed that our method achieved an average increase of nearly 8% in kappa score over EEGNet in decoding four-class MI tasks in 2D paradigms. Consistent performance across both VR and 2D paradigms confirmed the algorithm's effectiveness and applicability in multi-scenario MI decoding. This study introduces a novel immersive MI paradigm and decoding framework, offering a promising approach for enhancing user engagement in neurorehabilitation and advancing EEG-based intention recognition in VR environments.

Digital Behaviour Change Interventions (DBCIs) aim at improving individual health by engaging various means of Information and Communication Technology (ICT), including mobile apps and wearables. Participant intervention fatigue may happen when DBCIs become too frequent, repetitive, demanding, or lack perceived relevance, and this may result in participants' reduced motivation and adherence over time. Advancing technology-supported engagement mechanisms is therefore of utmost importance. To address this problem, we present a reference and solution architecture based on open-source technologies and open Application Programming Interfaces (Open APIs). First, we integrated a Large Language Model (LLM) component into the DBCI design. Second, to support context-awareness, we enhanced this integration by adding a Geographic Information Systems (GIS) element. Our pilot implemented AI4Motion platform targets both personalization and contextualization aspects of DBCIs. Our work contributes to the emerging discussion on LLM/GIS-related system design patterns for digital platforms supporting Ecological Momentary Assessment (EMA), Experience Sampling Method (ESM), and Just-in-Time Adaptive Interventions (JITAIs).