Utilizing large language models for detecting hospital-acquired conditions: an empirical study on pulmonary embolism.
利用大型語言模型檢測醫院獲得的病症:針對肺栓塞的實證研究。
J Am Med Inform Assoc 2025-03-19
Local large language models for privacy-preserving accelerated review of historic echocardiogram reports.
本地大型語言模型用於隱私保護加速審查歷史心臟超聲報告。
J Am Med Inform Assoc 2024-04-30
Large Language Models Improve the Identification of Emergency Department Visits for Symptomatic Kidney Stones.
大型語言模型改善了對有症狀腎結石的急診就診識別。
medRxiv 2024-08-30
Large language models can effectively extract stroke and reperfusion audit data from medical free-text discharge summaries.
大型語言模型能有效地從醫療自由文本出院摘要中提取中風和再灌注審核數據。
J Clin Neurosci 2024-09-21
Extraction of clinical data on major pulmonary diseases from unstructured radiologic reports using a large language model.
使用大型語言模型從非結構化放射學報告中提取主要肺部疾病的臨床數據。
PLoS One 2024-11-25
這項研究探討大型語言模型(LLMs)在從非結構化放射報告中提取臨床數據的效果,專注於七種肺部疾病。研究分析了1,800份報告,並使用Google Gemini Pro 1.0、OpenAI的GPT-3.5和GPT-4進行數據提取。結果顯示,所有模型的準確率都很高,特別是GPT-4的表現最佳,敏感性和特異性均達到優秀水準。這些結果顯示,LLMs,尤其是GPT-4,可能成為醫生進行病歷審查的有效替代方案,提升非結構化放射數據的提取能力。
PubMedDOI
Utility of a large language model for extraction of clinical findings from healthcare data following lung ablation: A feasibility study.
肺部消融後從醫療數據中提取臨床發現的大型語言模型的實用性:一項可行性研究。
J Vasc Interv Radiol 2024-12-11
Large language models improve the identification of emergency department visits for symptomatic kidney stones.
大型語言模型改善了對有症狀腎結石的急診就診識別。
Sci Rep 2025-01-28
Integrating large language models with human expertise for disease detection in electronic health records.
將大型語言模型與人類專業知識整合以進行電子健康紀錄中的疾病檢測。
Comput Biol Med 2025-04-08
Extracting Pulmonary Embolism Diagnoses From Radiology Impressions Using GPT-4o: Large Language Model Evaluation Study.
使用 GPT-4o 從放射學印象中提取肺栓塞診斷:大型語言模型評估研究。
JMIR Med Inform 2025-04-09
Evaluating Large Language Models in Cardiovascular Antithrombotic Care: Performance, Accuracy, and Implications for Clinical Practice.
心血管抗血栓治療中大型語言模型的評估:表現、準確性及其對臨床實務的影響
Can J Cardiol 2025-04-16
這項研究發現,Claude 3 Opus 在心血管抗凝治療案例的準確度勝過其他大型語言模型和臨床醫師,正確率達85%。部分LLMs表現媲美甚至超越有經驗醫師,但免費版模型有時會給出不佳或不安全的建議。所有LLMs在生活型態和飲食建議上表現穩定。研究提醒,醫療決策時應謹慎選用並驗證LLMs。
PubMedDOI