The advent of Large Language Models (LLMs) offers potential solutions to address problems such as shortage of medical resources and low diagnostic consistency in psychiatric clinical practice. Despite this potential, a robust and comprehensive benchmarking framework to assess the efficacy of LLMs in authentic psychiatric clinical environments is absent. This has impeded the advancement of specialized LLMs tailored to psychiatric applications. In response to this gap, by incorporating clinical demands in psychiatry and clinical data, we proposed a benchmarking system, PsychBench, to evaluate the practical performance of LLMs in psychiatric clinical settings. We conducted a comprehensive quantitative evaluation of 16 LLMs using PsychBench, and investigated the impact of prompt design, chain-of-thought reasoning, input text length, and domain-specific knowledge fine-tuning on model performance. Through detailed error analysis, we identified strengths and potential limitations of the existing models and suggested directions for improvement. Subsequently, a clinical reader study involving 60 psychiatrists of varying seniority was conducted to further explore the practical benefits of ex
Detecting arousals in sleep is essential for diagnosing sleep disorders. However, using Machine Learning (ML) in clinical practice is impeded by fundamental issues, primarily due to mismatches between clinical protocols and ML methods. Clinicians typically annotate only the onset of arousals, while ML methods rely on annotations for both the beginning and end. Additionally, there is no standardized evaluation methodology tailored to clinical needs for arousal detection models. This work addresses these issues by introducing a novel post-processing and evaluation framework emphasizing approximate localization and precise event count (ALPEC) of arousals. We recommend that ML practitioners focus on detecting arousal onsets, aligning with clinical practice. We examine the impact of this shift on current training and evaluation schemes, addressing simplifications and challenges. We utilize a novel comprehensive polysomnographic dataset (CPS) that reflects the aforementioned clinical annotation constraints and includes modalities not present in existing polysomnographic datasets. We release the dataset alongside this paper, demonstrating the benefits of leveraging multimodal data for aro
Bioinformatics platforms have significantly changed clinical diagnostics by facilitating the analysis of genomic data, thereby advancing personalized medicine and improving patient care. This study examines the integration, usage patterns, challenges, and impact of the Galaxy platform within clinical diagnostics laboratories. We employed a convergent parallel mixed-methods design, collecting quantitative survey data and qualitative insights from structured interviews with fifteen participants across various clinical roles. The findings indicate a wide adoption of Galaxy, with participants expressing high satisfaction due to its user-friendly interface and notable improvements in workflow efficiency and diagnostic accuracy. Challenges such as data security and training needs were also identified, highlighting the platform's role in simplifying complex data analysis tasks. This study contributes to understanding the transformative potential of Galaxy in clinical practice and offers recommendations for optimizing its integration and functionality. These insights are crucial for advancing clinical diagnostics and enhancing patient outcomes.
Recently natural language processing (NLP) tools have been developed to identify and extract salient risk indicators in electronic health records (EHRs). Sentiment analysis, although widely used in non-medical areas for improving decision making, has been studied minimally in the clinical setting. In this study, we undertook, to our knowledge, the first domain adaptation of sentiment analysis to psychiatric EHRs by defining psychiatric clinical sentiment, performing an annotation project, and evaluating multiple sentence-level sentiment machine learning (ML) models. Results indicate that off-the-shelf sentiment analysis tools fail in identifying clinically positive or negative polarity, and that the definition of clinical sentiment that we provide is learnable with relatively small amounts of training data. This project is an initial step towards further refining sentiment analysis methods for clinical use. Our long-term objective is to incorporate the results of this project as part of a machine learning model that predicts inpatient readmission risk. We hope that this work will initiate a discussion concerning domain adaptation of sentiment analysis to the clinical setting.
Using three years of the Journal Citation Reports (2011, 2012, and 2013), indicators of transitions in 2012 (between 2011 and 2013) are studied using methodologies based on entropy statistics. Changes can be indicated at the level of journals using the margin totals of entropy production along the row or column vectors, but also at the level of links among journals by importing the transition matrices into network analysis and visualization programs (and using community-finding algorithms). Seventy-four journals are flagged in terms of discontinuous changes in their citations; but 3,114 journals are involved in "hot" links. Most of these links are embedded in a main component; 78 clusters (containing 172 journals) are flagged as potential "hot spots" emerging at the network level. An additional finding is that PLoS ONE introduced a new communication dynamics into the database. The limitations of the methodology are elaborated using an example. The results of the study indicate where developments in the citation dynamics can be considered as significantly unexpected. This can be used as heuristic information; but what a "hot spot" in terms of the entropy statistics of aggregated cit
Large Language Models (LLMs) have demonstrated considerable potential in general practice. However, existing benchmarks and evaluation frameworks primarily depend on exam-style or simplified question-answer formats, lacking a competency-based structure aligned with the real-world clinical responsibilities encountered in general practice. Consequently, the extent to which LLMs can reliably fulfill the duties of general practitioners (GPs) remains uncertain. In this work, we propose a novel evaluation framework to assess the capability of LLMs to function as GPs. Based on this framework, we introduce a general practice benchmark (GPBench), whose data are meticulously annotated by domain experts in accordance with routine clinical practice standards. We evaluate ten state-of-the-art LLMs and analyze their competencies. Our findings indicate that current LLMs are not suitable for autonomous deployment in clinical general practice and that all realistic applications require continuous human oversight; further optimization specifically tailored to the daily responsibilities of GPs remains essential.
The competency of any intelligent agent is bounded by its formal account of the world in which it operates. Clinical AI lacks such an account. Existing frameworks address evaluation, regulation, or system design in isolation, without a shared model of the clinical world to connect them. We introduce the Clinical World Model, a framework that formalizes care as a tripartite interaction among Patient, Provider, and Ecosystem. To formalize how any agent, whether human or artificial, transforms information into clinical action, we develop parallel decision-making architectures for providers, patients, and AI agents, grounded in validated principles of clinical cognition. The Clinical AI Skill-Mix operationalizes competency through eight dimensions. Five define the clinical competency space (condition, phase, care setting, provider role, and task) and three specify how AI engages human reasoning (assigned authority, agent facing, and anchoring layer). The combinatorial product of these dimensions yields a space of billions of distinct competency coordinates. A central structural implication is that validation within one coordinate provides minimal evidence for performance in another, re
We introduce SoftTiger, a clinical large language model (CLaM) designed as a foundation model for healthcare workflows. The narrative and unstructured nature of clinical notes is a major obstacle for healthcare intelligentization. We address a critical problem of structuring clinical notes into clinical data, according to international interoperability standards. We collect and annotate data for three subtasks, namely, international patient summary, clinical impression and medical encounter. We then supervised fine-tuned a state-of-the-art LLM using public and credentialed clinical data. The training is orchestrated in a way that the target model can first support basic clinical tasks such as abbreviation expansion and temporal information extraction, and then learn to perform more complex downstream clinical tasks. Moreover, we address several modeling challenges in the healthcare context, e.g., extra long context window. Our blind pairwise evaluation shows that SoftTiger outperforms other popular open-source models and GPT-3.5, comparable to Gemini-pro, with a mild gap from GPT-4. We believe that LLMs may become a step-stone towards healthcare digitalization and democratization.
Empiric antibiotic prescribing in high-risk clinical contexts often requires decision making under conditions of incomplete information, where inappropriate coverage or unjustified escalation may compromise safety and antimicrobial stewardship. While clinical decision-support systems have been proposed to assist in this process, many approaches lack explicit governance and evaluation mechanisms defining scope, abstention conditions, recommendation permissibility, and expected system behavior. This work specifies a governance and evaluation framework for deterministic clinical decision-support systems operating under explicitly constrained scope. Deterministic behavior is adopted to ensure that identical inputs yield identical outputs, supporting transparency, auditability, and conservative decision support in high-risk prescribing contexts. The framework treats governance as a first-class design component, separating clinical decision logic from rule-based mechanisms that determine whether a recommendation may be issued. Explicit abstention, deterministic stewardship constraints, and exclusion rules are formalized as core constructs. The framework defines an evaluation methodology
Objectives: Electronic health records (EHRs) are only a first step in capturing and utilizing health-related data - the challenge is turning that data into useful information. Furthermore, EHRs are increasingly likely to include data relating to patient outcomes, functionality such as clinical decision support, and genetic information as well, and, as such, can be seen as repositories of increasingly valuable information about patients' health conditions and responses to treatment over time. Methods: We describe a case study of 423 patients treated by Centerstone within Tennessee and Indiana in which we utilized electronic health record data to generate predictive algorithms of individual patient treatment response. Multiple models were constructed using predictor variables derived from clinical, financial and geographic data. Results: For the 423 patients, 101 deteriorated, 223 improved and in 99 there was no change in clinical condition. Based on modeling of various clinical indicators at baseline, the highest accuracy in predicting individual patient response ranged from 70-72% within the models tested. In terms of individual predictors, the Centerstone Assessment of Recovery Le
We compare the network of aggregated journal-journal citation relations provided by the Journal Citation Reports (JCR) 2012 of the Science and Social Science Citation Indexes (SCI and SSCI) with similar data based on Scopus 2012. First, global maps were developed for the two sets separately; sets of documents can then be compared using overlays to both maps. Using fuzzy-string matching and ISSN numbers, we were able to match 10,524 journal names between the two sets; that is, 96.4% of the 10,936 journals contained in JCR or 51.2% of the 20,554 journals covered by Scopus. Network analysis was then pursued on the set of journals shared between the two databases and the two sets of unique journals. Citations among the shared journals are more comprehensively covered in JCR than Scopus, so the network in JCR is denser and more connected than in Scopus. The ranking of shared journals in terms of indegree (that is, numbers of citing journals) or total citations is similar in both databases overall (Spearman's \r{ho} > 0.97), but some individual journals rank very differently. Journals that are unique to Scopus seem to be less important--they are citing shared journals rather than bein
Introduction: Semantic search, which retrieves documents based on conceptual similarity rather than keywords, offers advantages for retrieval of clinical information. However, deploying semantic search across health systems, comprising hundreds of millions of clinical notes, presents formidable engineering, cost, and governance challenges that have prevented institutional adoption. Methods: We deployed a semantic search system at a large children's hospital indexing 166 million clinical notes (484 million embedding vectors) from 1.68 million patients. The system uses instruction-tuned qwen3-embedding-0.6B embeddings, stores vectors with storage-optimized indexing, maintains full-text metadata in a low-latency key-value store, and operates within a HIPAA-compliant governance framework. We evaluated the system by optimizing the model and chunking strategy using a physician-authored benchmark, characterizing full-scale performance (cost, latency, retrieval quality), and assessing clinical utility via chart abstraction efficiency and comparison to ICD-10 cohort generation. Results: The system delivers sub-second query latency with monthly operational costs of ~USD 4,000. Qwen3 embeddin
The increasing availability of unstructured clinical narratives in electronic health records (EHRs) has created new opportunities for automated disease characterization, cohort identification, and clinical decision support. However, modeling long, domain-specific clinical text remains challenging due to limited labeled data, severe class imbalance, and the high computational cost of adapting large pretrained language models. This study presents a GPT-based architecture for clinical text classification that adapts a pretrained decoder-only Transformer using a selective fine-tuning strategy. Rather than updating all model parameters, the majority of the GPT-2 backbone is frozen, and training is restricted to the final Transformer block, the final layer normalization, and a lightweight classification head. This approach substantially reduces the number of trainable parameters while preserving the representational capacity required to model complex clinical language. The proposed method is evaluated on radiology reports from the MIMIC-IV-Note dataset using uncertainty-aware CheXpert-style labels derived directly from report text. Experiments cover multiple problem formulations, includi
We evaluate the impact of large language model-based clinical decision support in live care. In partnership with Penda Health, a network of primary care clinics in Nairobi, Kenya, we studied AI Consult, a tool that serves as a safety net for clinicians by identifying potential documentation and clinical decision-making errors. AI Consult integrates into clinician workflows, activating only when needed and preserving clinician autonomy. We conducted a quality improvement study, comparing outcomes for 39,849 patient visits performed by clinicians with or without access to AI Consult across 15 clinics. Visits were rated by independent physicians to identify clinical errors. Clinicians with access to AI Consult made relatively fewer errors: 16% fewer diagnostic errors and 13% fewer treatment errors. In absolute terms, the introduction of AI Consult would avert diagnostic errors in 22,000 visits and treatment errors in 29,000 visits annually at Penda alone. In a survey of clinicians with AI Consult, all clinicians said that AI Consult improved the quality of care they delivered, with 75% saying the effect was "substantial". These results required a clinical workflow-aligned AI Consult i
Using "Analyze Results" at the Web of Science, one can directly generate overlays onto global journal maps of science. The maps are based on the 10,000+ journals contained in the Journal Citation Reports (JCR) of the Science and Social Science Citation Indices (2011). The disciplinary diversity of the retrieval is measured in terms of Rao-Stirling's "quadratic entropy." Since this indicator of interdisciplinarity is normalized between zero and one, the interdisciplinarity can be compared among document sets and across years, cited or citing. The colors used for the overlays are based on Blondel et al.'s (2008) community-finding algorithms operating on the relations journals included in JCRs. The results can be exported from VOSViewer with different options such as proportional labels, heat maps, or cluster density maps. The maps can also be web-started and/or animated (e.g., using PowerPoint). The "citing" dimension of the aggregated journal-journal citation matrix was found to provide a more comprehensive description than the matrix based on the cited archive. The relations between local and global maps and their different functions in studying the sciences in terms of journal lit
A number of journal classification systems have been developed in bibliometrics since the launch of the Citation Indices by the Institute of Scientific Information (ISI) in the 1960s. These systems are used to normalize citation counts with respect to field-specific citation patterns. The best known system is the so-called "Web-of-Science Subject Categories" (WCs). In other systems papers are classified by algorithmic solutions. Using the Journal Citation Reports 2014 of the Science Citation Index and the Social Science Citation Index (n of journals = 11,149), we examine options for developing a new system based on journal classifications into subject categories using aggregated journal-journal citation data. Combining routines in VOSviewer and Pajek, a tree-like classification is developed. At each level one can generate a map of science for all the journals subsumed under a category. Nine major fields are distinguished at the top level. Further decomposition of the social sciences is pursued for the sake of example with a focus on journals in information science (LIS) and science studies (STS). The new classification system improves on alternative options by avoiding the problem
This paper is dedicated to the design and evaluation of the first AMR parser tailored for clinical notes. Our objective was to facilitate the precise transformation of the clinical notes into structured AMR expressions, thereby enhancing the interpretability and usability of clinical text data at scale. Leveraging the colon cancer dataset from the Temporal Histories of Your Medical Events (THYME) corpus, we adapted a state-of-the-art AMR parser utilizing continuous training. Our approach incorporates data augmentation techniques to enhance the accuracy of AMR structure predictions. Notably, through this learning strategy, our parser achieved an impressive F1 score of 88% on the THYME corpus's colon cancer dataset. Moreover, our research delved into the efficacy of data required for domain adaptation within the realm of clinical notes, presenting domain adaptation data requirements for AMR parsing. This exploration not only underscores the parser's robust performance but also highlights its potential in facilitating a deeper understanding of clinical narratives through structured semantic representations.
Dyads of journals related by citations can agglomerate into specialties through the mechanism of triadic closure. Using the Journal Citation Reports 2011, 2012, and 2013, we analyze triad formation as indicators of integration (specialty growth) and disintegration (restructuring). The strongest integration is found among the large journals that report on studies in different scientific specialties, such as PLoS ONE, Nature Communications, Nature, and Science. This tendency towards large-scale integration has not yet stabilized. Using the Islands algorithm, we also distinguish 51 local maxima of integration. We zoom into the cited articles that carry the integration for: (i) a new development within high-energy physics and (ii) an emerging interface between the journals Applied Mathematical Modeling and the International Journal of Advanced Manufacturing Technology. In the first case, integration is brought about by a specific communication reaching across specialty boundaries, whereas in the second, the dyad of journals indicates an emerging interface between specialties. These results suggest that integration picks up substantive developments at the specialty level. An advantage o
This study examines the role of top-tier conference publications in Hungarian computer science research. We show that the national scientometric practice, which is currently journal-oriented, diverges from international norms, creating incentive distortions in researcher evaluation. By linking multiple databases (iCore, DBLP, MTMT, MTA-ATT), we mapped Hungarian-affiliated CORE A* and A conference papers, their temporal and thematic distribution, and author trajectories. Our results indicate that, in theoretical fields, publishing at international conferences became common earlier than in applied fields. At the same time, in applied fields, successful researchers are more likely to continue their careers in foreign institutions or in industry positions. Overall, a substantial share of the already established, internationally most successful researchers are now affiliated with institutions abroad. We recommend recognizing CORE A* papers as equivalent to D1 and CORE A papers as equivalent to Q1 journals in national evaluation systems.
In this paper we define Clinical Data Intelligence as the analysis of data generated in the clinical routine with the goal of improving patient care. We define a science of a Clinical Data Intelligence as a data analysis that permits the derivation of scientific, i.e., generalizable and reliable results. We argue that a science of a Clinical Data Intelligence is sensible in the context of a Big Data analysis, i.e., with data from many patients and with complete patient information. We discuss that Clinical Data Intelligence requires the joint efforts of knowledge engineering, information extraction (from textual and other unstructured data), and statistics and statistical machine learning. We describe some of our main results as conjectures and relate them to a recently funded research project involving two major German university hospitals.