Diagnosing our datasets: How does my language model learn clinical information?

• URL: http://arxiv.org/abs/2505.15024v2
• Date: Fri, 23 May 2025 01:09:11 GMT
• Title: Diagnosing our datasets: How does my language model learn clinical information?
• Abstract summary: Large language models (LLMs) have performed well across various clinical natural language processing tasks.<n>We examine how popular open-source LLMs learn clinical information from large mined corpora.
• Score: 7.099154613395202
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Large language models (LLMs) have performed well across various clinical natural language processing tasks, despite not being directly trained on electronic health record (EHR) data. In this work, we examine how popular open-source LLMs learn clinical information from large mined corpora through two crucial but understudied lenses: (1) their interpretation of clinical jargon, a foundational ability for understanding real-world clinical notes, and (2) their responses to unsupported medical claims. For both use cases, we investigate the frequency of relevant clinical information in their corresponding pretraining corpora, the relationship between pretraining data composition and model outputs, and the sources underlying this data. To isolate clinical jargon understanding, we evaluate LLMs on a new dataset MedLingo. Unsurprisingly, we find that the frequency of clinical jargon mentions across major pretraining corpora correlates with model performance. However, jargon frequently appearing in clinical notes often rarely appears in pretraining corpora, revealing a mismatch between available data and real-world usage. Similarly, we find that a non-negligible portion of documents support disputed claims that can then be parroted by models. Finally, we classified and analyzed the types of online sources in which clinical jargon and unsupported medical claims appear, with implications for future dataset composition.

Related papers

• Future Querying: Can LLMs Serve as Implicit Medical World Models? [3.3341114691426257]
  We introduce future querying, a paradigm that probes whether large language models (LLMs) can function as implicit medical world models.<n>Our framework operates on unstructured clinical documentation using endpoint-agnostic training.<n>We show that small, locally fine-tuned open-weight models can match or approach larger proprietary systems.
  arXiv  Detail & Related papers  (2026-08-24T13:41:12Z)
• MedArena: Comparing LLMs for Medicine-in-the-Wild Clinician Preferences [50.71326426975699]
  We present MedArena, an interactive evaluation platform for medical large language models (LLMs)<n>Given a clinician-provided query, MedArena presents responses from two randomly selected models and asks the user to select the preferred response.<n>Out of 1571 preferences collected across 12 LLMs up to November 1, 2025, Gemini 2.0 Flash Thinking, Gemini 2.5 Pro, and GPT-4o were the top three models by Bradley-Terry rating.
  arXiv  Detail & Related papers  (2026-03-13T22:30:26Z)
• A Federated and Parameter-Efficient Framework for Large Language Model Training in Medicine [59.78991974851707]
  Large language models (LLMs) have demonstrated strong performance on medical benchmarks, including question answering and diagnosis.<n>Most medical LLMs are trained on data from a single institution, which faces limitations in generalizability and safety in heterogeneous systems.<n>We introduce the model-agnostic and parameter-efficient federated learning framework for adapting LLMs to medical applications.
  arXiv  Detail & Related papers  (2026-01-29T18:48:21Z)
• Harnessing Large Language Models for Precision Querying and Retrieval-Augmented Knowledge Extraction in Clinical Data Science [3.4325249294405555]
  This study applies Large Language Models (LLMs) to two foundational Electronic Health Record (EHR) data science tasks.<n>We test the ability of LLMs to interact accurately with large structured datasets for analytics.<n>We present a flexible evaluation framework that automatically generates synthetic question and answer pairs tailored to the characteristics of each dataset or task.
  arXiv  Detail & Related papers  (2026-01-28T14:57:36Z)
• Simulating Viva Voce Examinations to Evaluate Clinical Reasoning in Large Language Models [51.91760712805404]
  We introduce VivaBench, a benchmark for evaluating sequential clinical reasoning in large language models (LLMs)<n>Our dataset consists of 1762 physician-curated clinical vignettes structured as interactive scenarios that simulate a (oral) examination in medical training.<n>Our analysis identified several failure modes that mirror common cognitive errors in clinical practice.
  arXiv  Detail & Related papers  (2025-10-11T16:24:35Z)
• Bringing CLIP to the Clinic: Dynamic Soft Labels and Negation-Aware Learning for Medical Analysis [0.9944647907864256]
  We propose a novel approach that integrates clinically-enhanced dynamic soft labels and medical graphical alignment.<n>Our approach is easily integrated into the medical CLIP training pipeline and achieves state-of-the-art performance.
  arXiv  Detail & Related papers  (2025-05-28T08:00:18Z)
• Rephrasing Electronic Health Records for Pretraining Clinical Language Models [0.09819964822292428]
  We rephrase existing clinical notes using LLMs to generate synthetic pretraining corpora.<n>We find that augmenting original clinical notes with synthetic corpora from different LLMs improves performances even at a small token budget.
  arXiv  Detail & Related papers  (2024-11-28T06:12:28Z)
• When Raw Data Prevails: Are Large Language Model Embeddings Effective in Numerical Data Representation for Medical Machine Learning Applications? [8.89829757177796]
  We examine the effectiveness of vector representations from last hidden states of Large Language Models for medical diagnostics and prognostics. We focus on instruction-tuned LLMs in a zero-shot setting to represent abnormal physiological data and evaluate their utilities as feature extractors. Although findings suggest the raw data features still prevails in medical ML tasks, zero-shot LLM embeddings demonstrate competitive results.
  arXiv  Detail & Related papers  (2024-08-15T03:56:40Z)
• Dr-LLaVA: Visual Instruction Tuning with Symbolic Clinical Grounding [53.629132242389716]
  Vision-Language Models (VLM) can support clinicians by analyzing medical images and engaging in natural language interactions. VLMs often exhibit "hallucinogenic" behavior, generating textual outputs not grounded in contextual multimodal information. We propose a new alignment algorithm that uses symbolic representations of clinical reasoning to ground VLMs in medical knowledge.
  arXiv  Detail & Related papers  (2024-05-29T23:19:28Z)
• Don't Ignore Dual Logic Ability of LLMs while Privatizing: A Data-Intensive Analysis in Medical Domain [19.46334739319516]
  We study how the dual logic ability of LLMs is affected during the privatization process in the medical domain. Our results indicate that incorporating general domain dual logic data into LLMs not only enhances LLMs' dual logic ability but also improves their accuracy.
  arXiv  Detail & Related papers  (2023-09-08T08:20:46Z)
• Development and validation of a natural language processing algorithm to pseudonymize documents in the context of a clinical data warehouse [53.797797404164946]
  The study highlights the difficulties faced in sharing tools and resources in this domain. We annotated a corpus of clinical documents according to 12 types of identifying entities. We build a hybrid system, merging the results of a deep learning model as well as manual rules.
  arXiv  Detail & Related papers  (2023-03-23T17:17:46Z)
• Do We Still Need Clinical Language Models? [15.023633270864675]
  We show that relatively small specialized clinical models substantially outperform all in-context learning approaches. We release the code and the models used under the PhysioNet Credentialed Health Data license and data use agreement.
  arXiv  Detail & Related papers  (2023-02-16T05:08:34Z)
• sEHR-CE: Language modelling of structured EHR data for efficient and generalizable patient cohort expansion [0.0]
  sEHR-CE is a novel framework based on transformers to enable integrated phenotyping and analyses of heterogeneous clinical datasets. We validate our approach using primary and secondary care data from the UK Biobank, a large-scale research study.
  arXiv  Detail & Related papers  (2022-11-30T16:00:43Z)
• Retrieval-Augmented and Knowledge-Grounded Language Models for Faithful Clinical Medicine [68.7814360102644]
  We propose the Re$3$Writer method with retrieval-augmented generation and knowledge-grounded reasoning. We demonstrate the effectiveness of our method in generating patient discharge instructions.
  arXiv  Detail & Related papers  (2022-10-23T16:34:39Z)
• Cross-Lingual Knowledge Transfer for Clinical Phenotyping [55.92262310716537]
  We investigate cross-lingual knowledge transfer strategies to execute this task for clinics that do not use the English language. We evaluate these strategies for a Greek and a Spanish clinic leveraging clinical notes from different clinical domains. Our results show that using multilingual data overall improves clinical phenotyping models and can compensate for data sparseness.
  arXiv  Detail & Related papers  (2022-08-03T08:33:21Z)
• Self-supervised Answer Retrieval on Clinical Notes [68.87777592015402]
  We introduce CAPR, a rule-based self-supervision objective for training Transformer language models for domain-specific passage matching. We apply our objective in four Transformer-based architectures: Contextual Document Vectors, Bi-, Poly- and Cross-encoders. We report that CAPR outperforms strong baselines in the retrieval of domain-specific passages and effectively generalizes across rule-based and human-labeled passages.
  arXiv  Detail & Related papers  (2021-08-02T10:42:52Z)
• Benchmarking Automated Clinical Language Simplification: Dataset, Algorithm, and Evaluation [48.87254340298189]
  We construct a new dataset named MedLane to support the development and evaluation of automated clinical language simplification approaches. We propose a new model called DECLARE that follows the human annotation procedure and achieves state-of-the-art performance.
  arXiv  Detail & Related papers  (2020-12-04T06:09:02Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.