225 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.
The study introduces DXA Agent, an agentic workflow that optimizes bone mineral density modeling by balancing predictive performance with measurement burden, demonstrating that cost-efficient models outperform conventional approaches in the UK Biobank while maintaining comparable results in NHANES.
This randomized mixed-methods study reveals that oncologists' trust and preference for AI-generated literature reviews depend more on report design features—such as conciseness, verifiable citations, and explicit uncertainty—than on accuracy alone, as evidenced by significantly lower utility ratings for evidence-graded reports compared to standard formats despite similar reference quality.
This study demonstrates that a computationally efficient, fine-tuned DistilBERT encoder is a more viable and cost-effective solution for detecting self-repairs in spontaneous speech for clinical screening than large generative LLMs, which vary significantly in performance and prompt sensitivity.
This study presents and validates ET-Emb, a novel multimodal artificial intelligence framework that integrates structured clinical data with unstructured text embeddings to accurately predict optimal personalized treatment strategies for hepatocellular carcinoma patients.
This paper demonstrates that offline reinforcement learning policies for ICU sepsis management maintain stable, action-sensitive decision-support signals under increasing out-of-distribution severity shifts, as evidenced by consistent model-predicted survival rates and improved physiological stabilization scores despite declining observed clinical outcomes.
This paper evaluates clinical concept extraction and evidence-bounded terminology linking through a multisite pilot study, demonstrating that while extraction performance varies significantly by matching criteria, the availability of retrieved evidence critically influences linking decisions and that initial unmatched terms often represent plausible existing concepts rather than true ontology novelties.
This study demonstrates that for Super Learner ensembles composed of tree-based algorithms, data preprocessing yields negligible improvements in predictive performance and oracle behavior across most clinical datasets, suggesting that such preprocessing is not universally necessary and should be guided by specific dataset characteristics.
This paper presents the HPP Personal Health Agent, a domain-specialized AI system that combines population-level phenotypic data, expert clinical tools, and declarative behavioral constraints to eliminate hallucinations and achieve high accuracy in metabolic health reporting, demonstrating that trustworthy medical AI requires a rigorous, eval-driven systems architecture rather than relying solely on general-purpose language models.
The paper introduces LaBERT, a transformer-based model that significantly outperforms baseline methods in forecasting future laboratory measurements from longitudinal electronic health records by effectively learning clinically meaningful treatment effects and enabling personalized clinical decision support.
This study utilizes LLM-based analysis of caregiver forum posts to reveal that burden, unmet needs, and system failures in dementia caregiving form interconnected ecosystems that vary by caregiver role and relationship, arguing for integrated support strategies rather than isolated interventions.