NPJ Digital Medicine最新文献_第9页

Quality optimization with low positive factual hallucination for the HPI in hyperthyroidism admission notes using multi agent LLM with RAG 多药LLM联合RAG对甲亢入院记录中HPI的质量优化及低阳性幻觉

IF 15.2 1区医学

NPJ Digital Medicine Pub Date : 2026-10-07 DOI: 10.1038/s41746-026-03321-x

Qingya Zeng, Huiyu Lan, Hang Lv, Jing Lin, Pengxiang Zhan, Binteng Cai, Yacong Yang, Ji Huang, Lan Lin, Yan Chen, Xiongfeng Chen, Wei Lin, Yanchao Tan, Gang Chen

{"title":"Quality optimization with low positive factual hallucination for the HPI in hyperthyroidism admission notes using multi agent LLM with RAG","authors":"Qingya Zeng, Huiyu Lan, Hang Lv, Jing Lin, Pengxiang Zhan, Binteng Cai, Yacong Yang, Ji Huang, Lan Lin, Yan Chen, Xiongfeng Chen, Wei Lin, Yanchao Tan, Gang Chen","doi":"10.1038/s41746-026-03321-x","DOIUrl":"https://doi.org/10.1038/s41746-026-03321-x","url":null,"abstract":"Admission note quality affects clinical safety and research reliability, yet resident drafts often suffer from missing information, unclear logic, and other issues, requiring time-consuming manual review. Current large language models (LLMs) focus on text generation or formatting correction but rarely optimize clinical logic or mitigate hallucinations in specialized notes. We developed HQ-ANR (High-Quality Admission Note Refinement), a hyperthyroidism-specific quality-control model that integrates multi-agent collaboration with diagnosis-driven retrieval-augmented generation (RAG) to optimize the history of present illness (HPI) in admission notes. Built on a general-purpose LLM (Qwen3-32B), HQ-ANR anchors generation to a dual-source knowledge base (guidelines and expert notes) and employs specialized agents for symptoms, examinations, and treatments. In a blinded evaluation of 30 resident-drafted hyperthyroidism notes, HQ-ANR significantly improved median Physician Documentation Quality Instrument (PDQI-9) scores (from 39.00 to 42.00) and reduced the hallucination rate (positive factual hallucinations only) to 6.67%, compared to 26.67% for a general LLM with chain-of-thought (CoT) prompting. Efficiency analysis showed HQ-ANR reduced physician refinement time by an average of 112.9 s per note (p < 0.001). These results demonstrate that augmenting a general LLM with external knowledge and a multi-agent framework can achieve reliable, in-depth optimization of clinical notes without expensive vertical training, offering a practical pathway toward intelligent, safe documentation quality control and reducing physician documentation burden.","PeriodicalId":19349,"journal":{"name":"NPJ Digital Medicine","volume":"18 1","pages":""},"PeriodicalIF":15.2,"publicationDate":"2026-10-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149087879","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Prospective deployment of multimodal AI grading for medical student OSCEs 多模态AI评分在医学生oses中的前瞻性部署

IF 15.2 1区医学

NPJ Digital Medicine Pub Date : 2026-10-07 DOI: 10.1038/s41746-026-03352-4

Huong-Tra Ngo,Ameer Hamza Shakur,Michael Holcomb,Shinyoung Kang,David Hein,Judah Gruen,Hunter Schuler,Philip Jarrett,Thomas Dalton,Krystle Campbell,Daniel Scott,Andrew R. Jamieson

引用次数: 0

AI-assisted telerehabilitation in early Parkinson’s disease: a multicenter, randomized, multi-arm comparative trial 人工智能辅助远程康复治疗早期帕金森病：一项多中心、随机、多组比较试验

IF 15.2 1区医学

NPJ Digital Medicine Pub Date : 2026-10-07 DOI: 10.1038/s41746-026-03339-1

Huimin Chen, Huijing Liu, Yueying Liu, Miaoxian Xie, Wenyi Kou, Xinxin Ma, Yunfei Long, Dongdong Wu, Wei Du, Jing He, Shuhua Li, Haibo Chen, Xue Bai, Kezhong Zhang, Jianping Huang, Yi Wang, Wen Su

{"title":"AI-assisted telerehabilitation in early Parkinson’s disease: a multicenter, randomized, multi-arm comparative trial","authors":"Huimin Chen, Huijing Liu, Yueying Liu, Miaoxian Xie, Wenyi Kou, Xinxin Ma, Yunfei Long, Dongdong Wu, Wei Du, Jing He, Shuhua Li, Haibo Chen, Xue Bai, Kezhong Zhang, Jianping Huang, Yi Wang, Wen Su","doi":"10.1038/s41746-026-03339-1","DOIUrl":"https://doi.org/10.1038/s41746-026-03339-1","url":null,"abstract":"This exploratory, multicenter, randomized, rater-blinded, multi-arm comparative trial aimed to evaluate the feasibility and exploratory clinical signals of an AI-facilitated telerehabilitation system across different intervention durations in early-stage Parkinson’s disease (PD). One-hundred and twenty PD patients (Hoehn-Yahr stage ≤2.5) were randomized in a 1:1:1 ratio into short-term (1 month), medium-term (2 months), and long-term (3 months) training groups (15–30 min, 5 days/week). The primary outcome was the change in MDS-UPDRS Part III score from baseline to 3 months. Secondary outcomes included change in cognition, mood, gait parameters, activity of daily living, and safety. Final analysis included 71 participants. No significant differences were observed between groups for the primary motor outcome at 3 months. Post hoc responder analysis showed that the 3-month group had higher proportions reaching the exploratory cognitive responder threshold compared with the 1-month group. Exploratory correlations between MoCA changes and motor parameter changes were weak and did not remain statistically significant after false discovery rate correction. No falls or adverse events were recorded throughout the trial. Because the primary endpoint was negative, attrition was high, no usual-care control was included, and the cognitive finding was limited to an exploratory post-hoc responder analysis, these findings do not permit causal inference and should be interpreted as hypothesis-generating, requiring further validation in larger, prespecified multi-arm comparative trials (Chinese Clinical Trial Registry Number: ChiCTR2500108323).","PeriodicalId":19349,"journal":{"name":"NPJ Digital Medicine","volume":"24 1","pages":""},"PeriodicalIF":15.2,"publicationDate":"2026-10-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149087696","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Digital health capacity building across health workforce and education sector in the WHO European region 世卫组织欧洲区域卫生人力和教育部门的数字卫生能力建设

IF 15.2 1区医学

NPJ Digital Medicine Pub Date : 2026-10-07 DOI: 10.1038/s41746-026-03337-3

Wiriya Mahikul, Mabel L. Prendergast, Gabriele K. Kerr, Sara Sousi, Benedict Hayhoe, Geva Greenfield, David Novillo-Ortiz, Josip Car, Azeem Majeed, Ana Luísa Neves

{"title":"Digital health capacity building across health workforce and education sector in the WHO European region","authors":"Wiriya Mahikul, Mabel L. Prendergast, Gabriele K. Kerr, Sara Sousi, Benedict Hayhoe, Geva Greenfield, David Novillo-Ortiz, Josip Car, Azeem Majeed, Ana Luísa Neves","doi":"10.1038/s41746-026-03337-3","DOIUrl":"https://doi.org/10.1038/s41746-026-03337-3","url":null,"abstract":"Disparities in digital capacity building across countries highlight the need for global upskilling. This study examined digital health capacity building across policy, workforce, and pre-service education domains across 53 countries of the WHO Europe Region. This was a cross-sectional ecological study using publicly available datasets. Descriptive and logistic regression analyses were performed. By 2022, 52.9% of countries had a digital health education strategy, 69.1% offered in-service training, and 59.6% provided certified student training. Countries demonstrating all three indicators tended to have higher literacy rates and strong R&D spending. A national digital health education strategy was associated with higher private or commercial funding for digital health, while more hospital beds per 1000 population were associated with lower odds of in-service digital health training. However, no factors were associated with health science students’ certified training, while public–private partnerships showed a non-significant positive association. Digital health capacity building across these indicators was uneven. Private-sector engagement and public–private funding and collaboration emerge as plausible correlates of policy on digital health capacity building. These findings suggest that policymakers should prioritize sustainable investment and cross-sector collaboration to strengthen digital health capacity building while further longitudinal and within-country studies are needed to confirm these associations.","PeriodicalId":19349,"journal":{"name":"NPJ Digital Medicine","volume":"404 1","pages":""},"PeriodicalIF":15.2,"publicationDate":"2026-10-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149087666","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

An examiner-conditioned AI second marker for VR OSCEs VR osce的考官条件AI第二标记

IF 15.2 1区医学

NPJ Digital Medicine Pub Date : 2026-10-07 DOI: 10.1038/s41746-026-03336-4

Harry Rogers,Sally Shiels,Ashley Tomlinson,James Thomas,Nathan Gauge,James Aylward,Sajan B. Patel,Angela Feixue Wang,Helen Higham,Alison Noble

{"title":"An examiner-conditioned AI second marker for VR OSCEs","authors":"Harry Rogers,Sally Shiels,Ashley Tomlinson,James Thomas,Nathan Gauge,James Aylward,Sajan B. Patel,Angela Feixue Wang,Helen Higham,Alison Noble","doi":"10.1038/s41746-026-03336-4","DOIUrl":"https://doi.org/10.1038/s41746-026-03336-4","url":null,"abstract":"Abstract Objective Structured Clinical Examinations (OSCEs) are typically scored by a single examiner per station, exposing results to hard-to-audit examiner variability. We developed the Rater-Aware Verification Network (RAVEN), a multimodal Artificial Intelligence (AI) system fusing egocentric video, examiner verbalisations and marks, and virtual reality (VR) action logs as an examiner-conditioned second marker for paediatric OSCEs (retrospective evaluation; 120 students, 442 ratings, eight domains). A confidence-gated hybrid improved agreement with leave-one-examiner-out consensus on the pass/fail decision (AC1: +6.2 percentage points; p = 0.0004) and domain scores (mean AC2: +3.0 points, five of eight significant), with the largest gains on borderline-fail cases (+16.3 points of concordance with consensus). Comparing AI-inferred, rubric-specified, and examiner-articulated criteria revealed implicit practices absent from marking guidelines. Agreement is measured against a panel-derived reference rather than an external ground truth; the system is intended as an audit and flagging aid for human adjudication rather than an autonomous decision-maker.","PeriodicalId":19349,"journal":{"name":"NPJ Digital Medicine","volume":"10 1","pages":""},"PeriodicalIF":15.2,"publicationDate":"2026-10-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149088742","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Universal CT representations from anatomy to disease phenotype through stage-wise pretraining 通过阶段预训练，从解剖学到疾病表型的通用CT表征

IF 15.2 1区医学

NPJ Digital Medicine Pub Date : 2026-10-07 DOI: 10.1038/s41746-026-03343-5

Yuheng Li,Yuan Gao,Haoyu Dong,Yuxiang Lai,Shansong Wang,Mojtaba Safari,James E. Baciak,Xiaofeng Yang

引用次数: 0

SPIRIT-CONSORT-ELM: element-level annotated dataset and large language model approach for assessing randomized controlled trial reporting spirit - consortium - elm：用于评估随机对照试验报告的元素级注释数据集和大型语言模型方法

IF 15.2 1区医学

NPJ Digital Medicine Pub Date : 2026-10-06 DOI: 10.1038/s41746-026-03318-6

Lan Jiang, Xiangji Ying, Andrew W. Brown, Mengfei Lan, Wenxuan Song, Joe D. Menke, Colby J. Vorland, Evan Mayo-Wilson, Halil Kilicoglu

{"title":"SPIRIT-CONSORT-ELM: element-level annotated dataset and large language model approach for assessing randomized controlled trial reporting","authors":"Lan Jiang, Xiangji Ying, Andrew W. Brown, Mengfei Lan, Wenxuan Song, Joe D. Menke, Colby J. Vorland, Evan Mayo-Wilson, Halil Kilicoglu","doi":"10.1038/s41746-026-03318-6","DOIUrl":"https://doi.org/10.1038/s41746-026-03318-6","url":null,"abstract":"Randomized controlled trials (RCTs) are central to assessing the benefits and harms of interventions, but incomplete reporting undermines their verifiability and usefulness. Although SPIRIT and CONSORT reporting guidelines promote complete reporting of RCT protocols and results publications, many RCTs remain incompletely reported. Automated manuscript checking could help improve reporting completeness before publication. We previously developed SPIRIT-CONSORT-TM, a corpus of 200 articles (100 protocol-results publication pairs) annotated with 83 checklist items from SPIRIT 2013 and CONSORT 2010, and trained models for item-level assessment. However, checklist items may comprise multiple constituent elements, which prior work did not capture or evaluate. Here, we extend the corpus with element-level annotations (SPIRIT-CONSORT-ELM) and formulate assessment as a machine reading comprehension task operationalized through 119 questions targeting specific reporting elements. Two annotators independently assessed 50 articles (25 pairs), with discrepancies resolved through discussion; one annotator assessed the remaining 150 articles. We then developed an automated pipeline combining PubMedBERT-based evidence retrieval with GPT-5-based question answering. Inter-annotator agreement was high (Gwet’s AC1: 0.782), and the pipeline achieved high performance (F1: 0.822, Gwet’s AC1: 0.796). Component analyses demonstrated the importance of evidence retrieval quality and modest benefits from illustrative in-context examples. SPIRIT-CONSORT-ELM provides a benchmark for fine-grained assessment of RCT reporting completeness, while the automated pipeline establishes a robust baseline and shows potential for supporting authors, reviewers, and editors.","PeriodicalId":19349,"journal":{"name":"NPJ Digital Medicine","volume":"41 1","pages":""},"PeriodicalIF":15.2,"publicationDate":"2026-10-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149087699","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Machine learning classification of digital exclusion in ageing populations across five international cohorts 机器学习对五个国际队列中老龄化人口数字排斥的分类

IF 15.2 1区医学

NPJ Digital Medicine Pub Date : 2026-10-06 DOI: 10.1038/s41746-026-03268-z

Shenwei Wan, Cong Fu, Can Lu, Zhiwen Ding, Shiyuan Wang, Lei Zhou, Xiaoguang Lei

引用次数: 0

LLM-derived narrative metrics in chronic pain: convergent validity and novel cognitive constructs 法学硕士衍生的慢性疼痛叙事指标：趋同效度和新的认知结构

IF 15.2 1区医学

NPJ Digital Medicine Pub Date : 2026-10-06 DOI: 10.1038/s41746-026-03359-x

Raquel Norel, Zhengwu Zhang, Jennifer Gewandter, Judith Naddour, Chadi G. Abdallah, Zhiyao Duan, Guillermo Cecchi, Paul Geha

引用次数: 0

Calibrated AI approach to pharmacovigilance using FAERS 使用FAERS进行药物警戒的校准人工智能方法