256 papers
Health informatics sits at the vibrant intersection of medicine, data science, and technology, transforming how we store, analyze, and utilize health information. This rapidly evolving field empowers clinicians and researchers to uncover patterns in patient data, improve diagnostic accuracy, and personalize treatment plans without getting lost in complex databases. By turning raw medical records into actionable insights, these innovations are reshaping the future of healthcare delivery and population health management.
At Gist.Science, we bridge the gap between cutting-edge research and public understanding by curating the latest preprints from medRxiv specifically within this domain. Our team processes every new submission in this category, providing both accessible plain-language explanations and detailed technical summaries to ensure the science is clear for everyone, from policymakers to curious readers. Below are the latest papers in health informatics, freshly distilled and ready for you to explore.
This study employed a human-centred co-design approach to develop and user-test an integrated heart failure dashboard that links patient-level clinical decision-making with service-level management, addressing the challenge of fragmented information across care settings.
This simulation study of six Pennsylvania counties demonstrates that the mortality reduction impact of increasing naloxone distribution and buprenorphine treatment varies significantly by county, driven by local baseline resources and historical dispensing patterns rather than uniform national trends.
This paper demonstrates that combining in-house and federated learning dashboards creates an optimal data quality assurance model for cross-institutional research, effectively balancing the thoroughness of local validation with the scalable, ecosystem-wide pattern detection of collaborative oversight.
This study demonstrates that the severity of depression and anxiety symptoms in a large cohort is significantly reflected in distinct physiological and behavioral metrics, such as sleep architecture, heart rate variability, and physical activity, captured by a consumer-grade smart ring.
The paper introduces SpyCAT, a novel semi-supervised, modeling-based anomaly detection system that leverages spatially context-aware transformers and vector quantization to effectively identify and localize subtle epileptogenic lesions in 3D brain MRI images, outperforming state-of-the-art methods.
This study presents a causal AI clinician trained on MIMIC and validated on eICU data that uses expert-grounded graphical models to generate optimal fluid and vasopressor recommendations for septic shock, demonstrating that adherence to these policies correlates with improved clinical outcomes while requiring significantly fewer variables than conventional predictive baselines.
This paper demonstrates that a hybrid retrieval architecture combining deterministic lexical matching with learned semantic search, query normalization, and rank fusion can enable these two paradigms to coexist safely over SNOMED CT without trade-offs, thereby improving clinical data entry accuracy for both precise terminology and ambiguous, abbreviated inputs while reducing the need for labor-intensive local vocabulary curation.
The paper introduces HemOncAgent, an AI system that integrates structured knowledge graphs and narrative care pathways to provide high-fidelity, hybrid retrieval of oncology knowledge, overcoming the limitations of single-source systems.
DualStream-MTCA is a hybrid deep learning model trained on MIMIC-IV data that effectively and simultaneously detects sepsis and heart failure in adult ICU patients using dual Bidirectional LSTM streams with cross-attention mechanisms, achieving high accuracy and strong generalizability in external validation.
This paper presents a curated resource of 7,763 significant real-world drug-diagnosis associations derived from UK and Danish electronic health records, which was manually annotated by medical experts to reveal that most observed co-occurrences do not represent direct treatment and to highlight the necessity of accounting for inter-annotator variability.