Mayo Clinic Proceedings. Digital health最新文献

英文中文

Artificial Intelligence Chest X-Ray Opportunistic Screening Model for Coronary Artery Calcium Deposition: A Multi-Objective Model With Multimodal Data Fusion 人工智能胸部x线筛查冠状动脉钙沉积的机会性模型：一个多目标多模态数据融合模型

Mayo Clinic Proceedings. Digital health

Pub Date : 2025-12-01 Epub Date: 2025-10-28 DOI: 10.1016/j.mcpdig.2025.100300

Jiwoong Jeong MS , Chieh-Ju Chao MD , Reza Arsanjani MD , Chadi Ayoub MBBS, PhD , Steven J. Lester MD , Milagros Pereyra MD , Ebram F. Said MD , Michael Roarke BS , Cecilia Tagle-Cornell MS , Laura M. Koepke MSN , Yi-Lin Tsai MD , Chen Jung-Hsuan MD , Chun-Chin Chang MD , Juan M. Farina MD , Hari Trivedi MD , Bhavik N. Patel MD, MBA , Imon Banerjee PhD

Objective

To create an opportunistic screening model to predict coronary calcium burden and associated cardiovascular risk using only commonly available frontal chest x-rays (CXR) and patient demographics.

Patients and Methods

We proposed a novel multitask learning framework and trained a model using 2121 patients with paired gated computed tomography scans and CXR images internally (Mayo Clinic) from January 1, 2012, to December 31, 2022, with coronary artery calcification (CAC) scores (0, 1-99, and 100+) as ground truths. Results from the internal training were validated on multiple external datasets (Emory University Healthcare and Taipei Veterans General Hospital—from January 1, 2012, to December 31, 2022) with significant racial and ethnic differences.

Results

Classification performance between 0, 1-99, and 100+ CAC scores performed moderately on both the internal test and external datasets, reaching average f1-scores of 0.71±0.04 for Mayo, 0.65±0.02 for Emory University Healthcare, and 0.70±0.06 for Taipei Veterans General Hospital. For the clinically relevant risk identification, the performance of our model on the internal and 2 external datasets reached area under the receiver operating curves of 0.86±0.02, 0.77±0.03, and 0.82±0.03 for 0 versus 400+, respectively. For 0 versus 100+, we achieved area under the receiver operating curve of 0.83±0.03, 0.71±0.02, and 0.78±0.01, respectively. Prospective evaluation across 3 Mayo Clinic sites is on par with the external validations and reports only minimal temporal drift.

Conclusion

Open-source fusion artificial intelligence-CXR model performed better than existing state-of-the-art models for predicting CAC scores only on internal cohort, with robust performance on external datasets. This proposed model may be useful as a robust, first-pass opportunistic screening method for cardiovascular risk from regular CXR.

目的建立一种机会性筛查模型，仅利用常用的胸部x光片（CXR）和患者人口统计学数据预测冠状动脉钙负荷和相关心血管风险。患者和方法我们提出了一个新的多任务学习框架，并使用2012年1月1日至2022年12月31日在梅奥诊所（Mayo Clinic）内部进行的2121例患者的配对门控制计算机断层扫描和CXR图像训练了一个模型，其中冠状动脉钙化（CAC）评分（0、1-99和100+）作为基本事实。内部训练的结果在多个外部数据集（Emory University Healthcare and Taipei Veterans General hospital，从2012年1月1日至2022年12月31日）上进行验证，具有显著的种族和民族差异。结果0、1 ~ 99、100+ CAC评分在内部和外部数据集的分类表现均为中等，梅奥医院的平均评分为0.71±0.04，埃默里大学医疗保健为0.65±0.02，台北退伍军人总医院为0.70±0.06。对于临床相关风险识别，我们的模型在内部和2个外部数据集上的表现在受试者工作曲线下分别达到0.86±0.02,0.77±0.03和0.82±0.03，分别为0和400+。对于0和100+，我们获得的受试者工作曲线下面积分别为0.83±0.03,0.71±0.02和0.78±0.01。3个Mayo诊所站点的前瞻性评估与外部验证相同，报告的时间偏差最小。结论开源融合人工智能- cxr模型仅在内部队列上预测CAC分数优于现有最先进的模型，在外部数据集上具有稳健的性能。该模型可作为常规CXR中心血管风险的一种稳健的第一次机会性筛查方法。

{"title":"Artificial Intelligence Chest X-Ray Opportunistic Screening Model for Coronary Artery Calcium Deposition: A Multi-Objective Model With Multimodal Data Fusion","authors":"Jiwoong Jeong MS , Chieh-Ju Chao MD , Reza Arsanjani MD , Chadi Ayoub MBBS, PhD , Steven J. Lester MD , Milagros Pereyra MD , Ebram F. Said MD , Michael Roarke BS , Cecilia Tagle-Cornell MS , Laura M. Koepke MSN , Yi-Lin Tsai MD , Chen Jung-Hsuan MD , Chun-Chin Chang MD , Juan M. Farina MD , Hari Trivedi MD , Bhavik N. Patel MD, MBA , Imon Banerjee PhD","doi":"10.1016/j.mcpdig.2025.100300","DOIUrl":"10.1016/j.mcpdig.2025.100300","url":null,"abstract":"<div><h3>Objective</h3><div>To create an opportunistic screening model to predict coronary calcium burden and associated cardiovascular risk using only commonly available frontal chest x-rays (CXR) and patient demographics.</div></div><div><h3>Patients and Methods</h3><div>We proposed a novel multitask learning framework and trained a model using 2121 patients with paired gated computed tomography scans and CXR images internally (Mayo Clinic) from January 1, 2012, to December 31, 2022, with coronary artery calcification (CAC) scores (0, 1-99, and 100+) as ground truths. Results from the internal training were validated on multiple external datasets (Emory University Healthcare and Taipei Veterans General Hospital—from January 1, 2012, to December 31, 2022) with significant racial and ethnic differences.</div></div><div><h3>Results</h3><div>Classification performance between 0, 1-99, and 100+ CAC scores performed moderately on both the internal test and external datasets, reaching average f1-scores of 0.71±0.04 for Mayo, 0.65±0.02 for Emory University Healthcare, and 0.70±0.06 for Taipei Veterans General Hospital. For the clinically relevant risk identification, the performance of our model on the internal and 2 external datasets reached area under the receiver operating curves of 0.86±0.02, 0.77±0.03, and 0.82±0.03 for 0 versus 400+, respectively. For 0 versus 100+, we achieved area under the receiver operating curve of 0.83±0.03, 0.71±0.02, and 0.78±0.01, respectively. Prospective evaluation across 3 Mayo Clinic sites is on par with the external validations and reports only minimal temporal drift.</div></div><div><h3>Conclusion</h3><div>Open-source fusion artificial intelligence-CXR model performed better than existing state-of-the-art models for predicting CAC scores only on internal cohort, with robust performance on external datasets. This proposed model may be useful as a robust, first-pass opportunistic screening method for cardiovascular risk from regular CXR.</div></div>","PeriodicalId":74127,"journal":{"name":"Mayo Clinic Proceedings. Digital health","volume":"3 4","pages":"Article 100300"},"PeriodicalIF":0.0,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145571780","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Increasing Retention in a Large-Scale Decentralized Clinical Trial: Learnings From the COVID-RED Trial 提高大规模分散临床试验的保留率：从COVID-RED试验中吸取的教训

Mayo Clinic Proceedings. Digital health

Pub Date : 2025-12-01 Epub Date: 2025-09-09 DOI: 10.1016/j.mcpdig.2025.100264

Laura C. Zwiers MPhil , Duco Veen PhD , Marianna Mitratza PhD , Timo B. Brakenhoff PhD , Brianna M. Goodale PhD , Paul Klaver MSc , Kay Y. Hage MSc , Marcel van Willigen PhD , George S. Downward PhD , Peter Lugtig PhD , Leendert van Maanen PhD , Stefan Van der Stigchel PhD , Peter van der Heijden PhD , Maureen Cronin PhD , Diederick E. Grobbee PhD , COVID-RED Consortium

Objective

To present retention strategies implemented in the coronavirus disease 2019 (COVID-19) rapid early detection trial, a decentralized trial investigating the use of a wearable device for severe acute respiratory syndrome coronavirus 2 detection, and to provide insights into study retention and investigate determinants of discontinuation.

Patients and Methods

The COVID-2019 rapid early detection trial collected data from 17,825 participants from February 22, 2021 to November 18, 2021. Participants wore a wearable device overnight and synchronized it with a mobile application on waking. Retention strategies included common and personalized activities. Multivariable logistic regression was used to identify participants at high risk of discontinuation after 6 months in the trial. Results were combined with insights from behavioral theory to target participants with additional telephone calls.

Results

Total of 14,326 (80.4%) participants remained in the trial after 6 months and 12,208 (68.5%) until the end of the trial. Multivariable logistic regression identified age, employment situation, living situation, and COVID-19 vaccination status as predictors of discontinuation. Subgroups at high risk of discontinuation were identified, and behavioral assessments indicated that the subgroup of vaccinated pensioners would receive additional telephone calls. Their dropout rate was 11.4% after telephone calls.

Conclusion

This study describes how innovative and targeted data-driven retention strategies can be applied in a large decentralized clinical trial and presents the implemented retention strategies and discontinuation rates. Results can serve as a starting point for designing retention strategies in future decentralized trials.

目的介绍2019冠状病毒病（COVID-19）快速早期检测试验（一项调查使用可穿戴设备检测严重急性呼吸综合征冠状病毒2的分散试验）中实施的保留策略，为研究保留提供见解，并调查终止的决定因素。患者和方法2019冠状病毒病快速早期检测试验于2021年2月22日至2021年11月18日收集了17825名参与者的数据。参与者在夜间佩戴可穿戴设备，并在醒来时将其与移动应用程序同步。留存策略包括普通活动和个性化活动。多变量逻辑回归用于确定试验6个月后停药风险高的参与者。结果与行为理论的见解相结合，向目标参与者提供额外的电话。结果6个月后，共有14326人（80.4%）仍在试验中，12208人（68.5%）直到试验结束。多变量logistic回归发现，年龄、就业状况、生活状况和COVID-19疫苗接种状况是停药的预测因素。确定了停止接种的高风险亚组，行为评估表明，接种疫苗的养恤金领取者亚组将接到额外的电话。通过电话后，他们的辍学率为11.4%。本研究描述了创新和有针对性的数据驱动的保留策略如何应用于大型分散临床试验，并介绍了实施的保留策略和停药率。结果可以作为设计未来分散试验中留存策略的起点。

{"title":"Increasing Retention in a Large-Scale Decentralized Clinical Trial: Learnings From the COVID-RED Trial","authors":"Laura C. Zwiers MPhil , Duco Veen PhD , Marianna Mitratza PhD , Timo B. Brakenhoff PhD , Brianna M. Goodale PhD , Paul Klaver MSc , Kay Y. Hage MSc , Marcel van Willigen PhD , George S. Downward PhD , Peter Lugtig PhD , Leendert van Maanen PhD , Stefan Van der Stigchel PhD , Peter van der Heijden PhD , Maureen Cronin PhD , Diederick E. Grobbee PhD , COVID-RED Consortium","doi":"10.1016/j.mcpdig.2025.100264","DOIUrl":"10.1016/j.mcpdig.2025.100264","url":null,"abstract":"<div><h3>Objective</h3><div>To present retention strategies implemented in the coronavirus disease 2019 (COVID-19) rapid early detection trial, a decentralized trial investigating the use of a wearable device for severe acute respiratory syndrome coronavirus 2 detection, and to provide insights into study retention and investigate determinants of discontinuation.</div></div><div><h3>Patients and Methods</h3><div>The COVID-2019 rapid early detection trial collected data from 17,825 participants from February 22, 2021 to November 18, 2021. Participants wore a wearable device overnight and synchronized it with a mobile application on waking. Retention strategies included common and personalized activities. Multivariable logistic regression was used to identify participants at high risk of discontinuation after 6 months in the trial. Results were combined with insights from behavioral theory to target participants with additional telephone calls.</div></div><div><h3>Results</h3><div>Total of 14,326 (80.4%) participants remained in the trial after 6 months and 12,208 (68.5%) until the end of the trial. Multivariable logistic regression identified age, employment situation, living situation, and COVID-19 vaccination status as predictors of discontinuation. Subgroups at high risk of discontinuation were identified, and behavioral assessments indicated that the subgroup of vaccinated pensioners would receive additional telephone calls. Their dropout rate was 11.4% after telephone calls.</div></div><div><h3>Conclusion</h3><div>This study describes how innovative and targeted data-driven retention strategies can be applied in a large decentralized clinical trial and presents the implemented retention strategies and discontinuation rates. Results can serve as a starting point for designing retention strategies in future decentralized trials.</div></div>","PeriodicalId":74127,"journal":{"name":"Mayo Clinic Proceedings. Digital health","volume":"3 4","pages":"Article 100264"},"PeriodicalIF":0.0,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145222262","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Correction Notices 调整通知

Mayo Clinic Proceedings. Digital health

Pub Date : 2025-12-01 Epub Date: 2025-11-07 DOI: 10.1016/j.mcpdig.2025.100305

引用次数: 0

Integrating U-Net in a LLM Supervisor Agent Pipeline for Pancreatic Ductal Adenocarcinoma Diagnosis 整合U-Net在LLM监督代理管道中诊断胰腺导管腺癌

Mayo Clinic Proceedings. Digital health

Pub Date : 2025-12-01 Epub Date: 2025-12-16 DOI: 10.1016/j.mcpdig.2025.100287

Rahul Gomes PhD, William L. Jerome BS, Sushil K. Garg MBBS

引用次数: 0

Graph-Based Deep Ensemble Learning to Enhance Diagnostic Efficiency in Lung Adenocarcinoma H&E-Stained Histopathological Subtyping 基于图的深度集成学习提高肺腺癌h&e染色病理分型的诊断效率

Mayo Clinic Proceedings. Digital health

Pub Date : 2025-12-01 Epub Date: 2025-12-16 DOI: 10.1016/j.mcpdig.2025.100285

Mohammad Mehdi Hosseini , Meghdad Sabouri Rad , Junze (Vincent) Huang , Rakesh Choudhary , Saverio J. Carello , Ola El-Zammar , Michel Nasr , Bardia Rodd

引用次数: 0

Exploring Novel Data-Driven Clustering Methods for Uncovering Patterns in Longitudinal Neonatal Postoperative Temperature Measurements 探索新的数据驱动的聚类方法揭示模式的纵向新生儿术后体温测量

Mayo Clinic Proceedings. Digital health

Pub Date : 2025-12-01 Epub Date: 2025-09-25 DOI: 10.1016/j.mcpdig.2025.100270

Stephanie M. Helman PhD , Nathan T. Riek PhD , Susan M. Sereika PhD , Ahmad P. Tafti PhD , Robert Olsen BS , J. William Gaynor MD , Amy Jo Lisanti PhD , Salah S. Al-Zaiti PhD

Objective

To identify distinct postoperative temperature trajectories in neonates with congenital heart defects after cardiopulmonary bypass (CPB), using advanced unsupervised machine learning clustering methods, corroborate findings, and evaluate their prognostic value on outcomes.

Patients and Methods

A secondary cohort analysis of prospective data collected from a single pediatric referral center’s CardioAccess data registry, consistent of neonates who underwent CPB between January 1, 2015, and January 1, 2019, was performed. Postoperative temperatures were extracted from medical records (48 hours). Group-based trajectory modeling (GBTM) performance was compared with self-organizing maps (SOM) and k-means clustering. Cluster membership and model fit were optimized for 3 temperature clusters per method. The primary outcome was a composite of postoperative complications. Clustering techniques were compared and associated with outcomes using adjusted multivariable binary logistic regression.

Results

Neonates of ≥34 weeks’ gestation underwent CPB (N=450). GBTM, SOM, and k-means identified membership for 3 groups: (1) persistent hypothermia (n=38 [9%]; n=49 [11%]; and n=40 [9%], respectively); (2) resolving hypothermia (n=233 [51%]; n=227 [50%]; and n=147 [33%], respectively); and (3) normothermia (n=179 [40%]; n=174 [39%]; and n=263 [58%], respectively). Concordance between techniques found strong agreement between GBTM and SOM (κ=0.92) and weak agreement between GBTM and k-means (κ=0.41). After adjustment, persistently hypothermic neonates compared with normothermic neonates were associated with higher odds of the complication composite outcome in the GBTM (odds ratio [OR], 2.8; 95% CI, 1.0-7.3; P=.04) and SOM (OR, 2.3; 95% CI, 1.0-5.4; P=.04) models, but not in the k-means model (OR, 1.4; 95% CI, 0.7-3.1; P=.38).

Conclusion

Exploring concordance between different machine learning techniques shows that temperature in neonates after CPB follows unique trajectories. Those exhibiting persistent hypothermia trends are at higher risk of adverse outcomes.

目的利用先进的无监督机器学习聚类方法，识别先天性心脏缺陷新生儿体外循环（CPB）术后不同的温度轨迹，验证结果，并评估其对预后的预测价值。患者和方法：对2015年1月1日至2019年1月1日期间接受CPB的新生儿的CardioAccess数据注册表收集的前瞻性数据进行二级队列分析。从医疗记录中提取术后48小时的温度。将基于组的轨迹建模（GBTM）与自组织映射（SOM）和k-means聚类进行了性能比较。每种方法对3个温度簇进行了聚类隶属度和模型拟合优化。主要结局是术后并发症的综合。采用调整后的多变量二元逻辑回归比较聚类技术并将其与结果相关联。结果≥34孕周新生儿行CPB （N=450）。GBTM、SOM和k-means识别出3组患者：(1)持续低温（n=38[9%]、n=49[11%]和n=40 [9%]）；(2)解决低温症（n=233 [51%], n=227 [50%], n=147 [33%]）；(3)正常母性贫血（n=179 [40%], n=174 [39%], n=263[58%]）。结果表明，GBTM与SOM之间的一致性较强（κ=0.92），而与k-means之间的一致性较弱（κ=0.41）。调整后，在GBTM模型中（比值比[OR]， 2.8; 95% CI, 1.0-7.3； P= 0.04）和SOM模型中（比值比[OR]， 2.3; 95% CI, 1.0-5.4； P= 0.04），持续低温新生儿与常温新生儿相比，并发症综合结局的几率更高，但在k-means模型中没有相关（比值比[OR]， 1.4; 95% CI, 0.7-3.1； P= 0.38）。结论探索不同机器学习技术之间的一致性表明CPB后新生儿的温度遵循独特的轨迹。那些表现出持续低体温趋势的人有更高的不良后果风险。

{"title":"Exploring Novel Data-Driven Clustering Methods for Uncovering Patterns in Longitudinal Neonatal Postoperative Temperature Measurements","authors":"Stephanie M. Helman PhD , Nathan T. Riek PhD , Susan M. Sereika PhD , Ahmad P. Tafti PhD , Robert Olsen BS , J. William Gaynor MD , Amy Jo Lisanti PhD , Salah S. Al-Zaiti PhD","doi":"10.1016/j.mcpdig.2025.100270","DOIUrl":"10.1016/j.mcpdig.2025.100270","url":null,"abstract":"<div><h3>Objective</h3><div>To identify distinct postoperative temperature trajectories in neonates with congenital heart defects after cardiopulmonary bypass (CPB), using advanced unsupervised machine learning clustering methods, corroborate findings, and evaluate their prognostic value on outcomes.</div></div><div><h3>Patients and Methods</h3><div>A secondary cohort analysis of prospective data collected from a single pediatric referral center’s CardioAccess data registry, consistent of neonates who underwent CPB between January 1, 2015, and January 1, 2019, was performed. Postoperative temperatures were extracted from medical records (48 hours). Group-based trajectory modeling (GBTM) performance was compared with self-organizing maps (SOM) and k-means clustering. Cluster membership and model fit were optimized for 3 temperature clusters per method. The primary outcome was a composite of postoperative complications. Clustering techniques were compared and associated with outcomes using adjusted multivariable binary logistic regression.</div></div><div><h3>Results</h3><div>Neonates of ≥34 weeks’ gestation underwent CPB (<em>N</em>=450). GBTM, SOM, and k-means identified membership for 3 groups: (1) persistent hypothermia (n=38 [9%]; n=49 [11%]; and n=40 [9%], respectively); (2) resolving hypothermia (n=233 [51%]; n=227 [50%]; and n=147 [33%], respectively); and (3) normothermia (n=179 [40%]; n=174 [39%]; and n=263 [58%], respectively). Concordance between techniques found strong agreement between GBTM and SOM (κ=0.92) and weak agreement between GBTM and k-means (κ=0.41). After adjustment, persistently hypothermic neonates compared with normothermic neonates were associated with higher odds of the complication composite outcome in the GBTM (odds ratio [OR], 2.8; 95% CI, 1.0-7.3; <em>P</em>=.04) and SOM (OR, 2.3; 95% CI, 1.0-5.4; <em>P</em>=.04) models, but not in the k-means model (OR, 1.4; 95% CI, 0.7-3.1; <em>P</em>=.38).</div></div><div><h3>Conclusion</h3><div>Exploring concordance between different machine learning techniques shows that temperature in neonates after CPB follows unique trajectories. Those exhibiting persistent hypothermia trends are at higher risk of adverse outcomes.</div></div>","PeriodicalId":74127,"journal":{"name":"Mayo Clinic Proceedings. Digital health","volume":"3 4","pages":"Article 100270"},"PeriodicalIF":0.0,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145417123","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A Technology Selection Tool Applying Multiple Criteria Decision Analysis for Virtual Care Implementation 应用多标准决策分析的虚拟护理实施技术选择工具

Mayo Clinic Proceedings. Digital health

Pub Date : 2025-12-01 Epub Date: 2025-09-03 DOI: 10.1016/j.mcpdig.2025.100260

Joseph P. Deason MBA , Scott J. Adams MD, PhD, FRCPC , Ahmad Rahman MSc , Stacey Lovo PhD , Ivar Mendez MD, PhD, FRCSC

Objective

To develop and pilot a technology selection tool (TST) designed to evaluate and recommend virtual care technologies tailored to specific community clinical needs.

Patients and Methods

Developed through collaborations among clinicians, software developers, technology experts, and health administrators, the TST uses a multiple criteria decision analysis framework to recommend technologies based on clinical relevance and technical quality. Its functionality was tested in a pilot project that assessed 5 technologies for their application in virtual wound care to support a remote community in Saskatchewan, Canada. The pilot study was completed March 7, 2025, through July 28, 2025.

Results

The TST identified the TeleVU Glass View as the optimal technology for virtual wound care. The TST generated product scores for the TeleVU Glass View (71.67), Teladoc Xpress (70.10), 19 Labs GALE (50.67), and TytoCare TytoKit (47.00), whereas disqualifying the Teladoc Lite Cart for not meeting the pass–fail portability criterion. TeleVU’s high product score resulted primarily from its technological attribute quality scores for Telestration (10), Audio (9), Video (9), and Share Content (9), which were all determined as clinically relevant for virtual wound care. The pilot enabled real-time wound care support by connecting local clinicians with virtual teams.

Conclusion

The TST offers a practical and adaptable tool to support evidence-based decision making for selecting technologies for specific clinical applications.

目的开发和试验一种技术选择工具（TST），旨在评估和推荐适合特定社区临床需求的虚拟护理技术。患者和方法通过临床医生、软件开发人员、技术专家和卫生管理人员之间的合作开发，TST使用多标准决策分析框架，根据临床相关性和技术质量推荐技术。它的功能在一个试点项目中进行了测试，该项目评估了5种技术在虚拟伤口护理中的应用，以支持加拿大萨斯喀彻温省的一个偏远社区。试点研究于2025年3月7日至2025年7月28日完成。结果TST认为TeleVU Glass View是虚拟创面护理的最佳技术。TST为TeleVU Glass View（71.67）、Teladoc Xpress（70.10）、19 Labs GALE（50.67）和TytoCare TytoKit（47.00）生成了产品分数，而Teladoc Lite Cart因不符合通过-失败可移植性标准而被取消资格。TeleVU的高产品得分主要来自Telestration（10分）、Audio（9分）、Video（9分）和Share Content（9分）的技术属性质量得分，这些都被认为与虚拟伤口护理具有临床相关性。该试点通过将当地临床医生与虚拟团队联系起来，实现了实时伤口护理支持。结论TST是一种实用且适应性强的工具，可为临床特定应用的技术选择提供循证决策支持。

{"title":"A Technology Selection Tool Applying Multiple Criteria Decision Analysis for Virtual Care Implementation","authors":"Joseph P. Deason MBA , Scott J. Adams MD, PhD, FRCPC , Ahmad Rahman MSc , Stacey Lovo PhD , Ivar Mendez MD, PhD, FRCSC","doi":"10.1016/j.mcpdig.2025.100260","DOIUrl":"10.1016/j.mcpdig.2025.100260","url":null,"abstract":"<div><h3>Objective</h3><div>To develop and pilot a technology selection tool (TST) designed to evaluate and recommend virtual care technologies tailored to specific community clinical needs.</div></div><div><h3>Patients and Methods</h3><div>Developed through collaborations among clinicians, software developers, technology experts, and health administrators, the TST uses a multiple criteria decision analysis framework to recommend technologies based on clinical relevance and technical quality. Its functionality was tested in a pilot project that assessed 5 technologies for their application in virtual wound care to support a remote community in Saskatchewan, Canada. The pilot study was completed March 7, 2025, through July 28, 2025.</div></div><div><h3>Results</h3><div>The TST identified the TeleVU Glass View as the optimal technology for virtual wound care. The TST generated product scores for the TeleVU Glass View (71.67), Teladoc Xpress (70.10), 19 Labs GALE (50.67), and TytoCare TytoKit (47.00), whereas disqualifying the Teladoc Lite Cart for not meeting the pass–fail portability criterion. TeleVU’s high product score resulted primarily from its technological attribute quality scores for Telestration (10), Audio (9), Video (9), and Share Content (9), which were all determined as clinically relevant for virtual wound care. The pilot enabled real-time wound care support by connecting local clinicians with virtual teams.</div></div><div><h3>Conclusion</h3><div>The TST offers a practical and adaptable tool to support evidence-based decision making for selecting technologies for specific clinical applications.</div></div>","PeriodicalId":74127,"journal":{"name":"Mayo Clinic Proceedings. Digital health","volume":"3 4","pages":"Article 100260"},"PeriodicalIF":0.0,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145268892","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Identifying Bias at Scale in Clinical Notes Using Large Language Models 使用大型语言模型识别临床笔记的尺度偏差

Mayo Clinic Proceedings. Digital health

Pub Date : 2025-12-01 Epub Date: 2025-10-14 DOI: 10.1016/j.mcpdig.2025.100296

Donald U. Apakama MD, MS , Kim-Anh-Nhi Nguyen MS , Daphnee Hyppolite MPA, RHIA , Shelly Soffer MD , Aya Mudrik BS , Emilia Ling MD, MBA, MS , Akini Moses MD , Ivanka Temnycky MS , Allison Glasser MBA , Rebecca Anderson MPH , Prathamesh Parchure MS , Evajoyce Woullard MS , Masoud Edalati PhD , Lili Chan MD, MS , Clair Kronk PhD , Robert Freeman RN , Arash Kia MD , Prem Timsina MD, PhD , Matthew A. Levin MD , Rohan Khera MD, MS , Girish N. Nadkarni MD, MPH

Objective

To evaluate whether generative pretrained transformer (GPT)-4 can detect and revise biased language in emergency department (ED) notes, against human-adjudicated gold-standard labels, and to identify modifiable factors associated with biased documentation.

Patients and Methods

We randomly sampled 50,000 ED medical and nursing notes from the Mount Sinai Health System (January 1, 2023, to December 31, 2023). We also randomly sampled 500 discharge notes from the Medical Information Mart for Intensive Care IV database. The GPT-4 flagged 4 types of bias: discrediting, stigmatizing/labeling, judgmental, and stereotyping. Two human reviewers verified model detections. We used multivariable logistic regression to examine associations between bias and health care utilization, presenting problems (eg, substance use), shift timing, and provider type. We then asked physicians to rate GPT-4’s proposed language revisions on a 10-point scale.

Results

The GPT-4 showed 97.6% sensitivity and 85.7% specificity compared with the human review. Biased language appeared in 6.5% (3229 of 50,000) of Mount Sinai notes and 7.4% (37 of 500) of Medical Information Mart for Intensive Care IV notes. In adjusted models, frequent health care utilization (adjusted odds ratio [aOR], 2.85; 95% CI, 1.95-4.17), substance use presentations (aOR, 3.09; 95% CI, 2.51-3.80), and overnight shifts (aOR, 1.37; 95% CI, 1.23-1.52) showed elevated odds of biased documentation. Physicians were more likely to include bias than nurses (aOR, 2.26; 95% CI, 2.07-2.46); GPT-4’s recommended revisions received mean physician ratings above 9 of 10.

Conclusion

The study showed that GPT-4 accurately detects biased language in clinical notes, identifies modifiable contributors to that bias, and delivers physician-endorsed revisions. This approach may help mitigate documentation bias and reduce disparities in care.

目的评估生成式预训练转换器(GPT)-4是否可以检测和修改急诊科（ED）笔记中的偏见语言，以对抗人类判定的金标准标签，并识别与偏见文件相关的可修改因素。患者与方法我们从西奈山卫生系统（2023年1月1日至2023年12月31日）随机抽取5万份急诊科医疗护理记录。我们还从重症监护医疗信息市场IV数据库中随机抽取500份出院记录。GPT-4标记了4种类型的偏见：诋毁，污名化/标签，判断和刻板印象。两名人工审查员验证了模型检测。我们使用多变量逻辑回归来检验偏倚与医疗保健利用、呈现问题（如物质使用）、轮班时间和提供者类型之间的关系。然后，我们要求医生对GPT-4提出的语言修订进行10分制的评分。结果GPT-4的敏感性为97.6%，特异性为85.7%。西奈山病历中有6.5%（3229 / 5万）存在语言偏差，重症监护IV期医疗信息市场病历中有7.4%（37 / 500）存在语言偏差。在调整后的模型中，频繁的医疗保健使用（调整优势比[aOR]， 2.85; 95% CI, 1.95-4.17）、物质使用表现（aOR, 3.09; 95% CI, 2.51-3.80）和夜班（aOR, 1.37; 95% CI, 1.23-1.52）显示有偏倚文献的几率升高。医生比护士更容易纳入偏倚（aOR, 2.26; 95% CI, 2.07-2.46）；GPT-4推荐的修订获得了医生平均9分以上的评分（满分10分）。该研究表明，GPT-4能准确地检测临床记录中的偏见语言，识别导致偏见的可修改因素，并提供医生认可的修订。这种方法可能有助于减轻文献偏倚和减少护理差异。

{"title":"Identifying Bias at Scale in Clinical Notes Using Large Language Models","authors":"Donald U. Apakama MD, MS , Kim-Anh-Nhi Nguyen MS , Daphnee Hyppolite MPA, RHIA , Shelly Soffer MD , Aya Mudrik BS , Emilia Ling MD, MBA, MS , Akini Moses MD , Ivanka Temnycky MS , Allison Glasser MBA , Rebecca Anderson MPH , Prathamesh Parchure MS , Evajoyce Woullard MS , Masoud Edalati PhD , Lili Chan MD, MS , Clair Kronk PhD , Robert Freeman RN , Arash Kia MD , Prem Timsina MD, PhD , Matthew A. Levin MD , Rohan Khera MD, MS , Girish N. Nadkarni MD, MPH","doi":"10.1016/j.mcpdig.2025.100296","DOIUrl":"10.1016/j.mcpdig.2025.100296","url":null,"abstract":"<div><h3>Objective</h3><div>To evaluate whether generative pretrained transformer (GPT)-4 can detect and revise biased language in emergency department (ED) notes, against human-adjudicated gold-standard labels, and to identify modifiable factors associated with biased documentation.</div></div><div><h3>Patients and Methods</h3><div>We randomly sampled 50,000 ED medical and nursing notes from the Mount Sinai Health System (January 1, 2023, to December 31, 2023). We also randomly sampled 500 discharge notes from the Medical Information Mart for Intensive Care IV database. The GPT-4 flagged 4 types of bias: discrediting, stigmatizing/labeling, judgmental, and stereotyping. Two human reviewers verified model detections. We used multivariable logistic regression to examine associations between bias and health care utilization, presenting problems (eg, substance use), shift timing, and provider type. We then asked physicians to rate GPT-4’s proposed language revisions on a 10-point scale.</div></div><div><h3>Results</h3><div>The GPT-4 showed 97.6% sensitivity and 85.7% specificity compared with the human review. Biased language appeared in 6.5% (3229 of 50,000) of Mount Sinai notes and 7.4% (37 of 500) of Medical Information Mart for Intensive Care IV notes. In adjusted models, frequent health care utilization (adjusted odds ratio [aOR], 2.85; 95% CI, 1.95-4.17), substance use presentations (aOR, 3.09; 95% CI, 2.51-3.80), and overnight shifts (aOR, 1.37; 95% CI, 1.23-1.52) showed elevated odds of biased documentation. Physicians were more likely to include bias than nurses (aOR, 2.26; 95% CI, 2.07-2.46); GPT-4’s recommended revisions received mean physician ratings above 9 of 10.</div></div><div><h3>Conclusion</h3><div>The study showed that GPT-4 accurately detects biased language in clinical notes, identifies modifiable contributors to that bias, and delivers physician-endorsed revisions. This approach may help mitigate documentation bias and reduce disparities in care.</div></div>","PeriodicalId":74127,"journal":{"name":"Mayo Clinic Proceedings. Digital health","volume":"3 4","pages":"Article 100296"},"PeriodicalIF":0.0,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145571261","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Foundation Models and Their Applications in Gastrointestinal Endoscopy 基础模型及其在胃肠内镜检查中的应用

Mayo Clinic Proceedings. Digital health

Pub Date : 2025-12-01 Epub Date: 2025-12-16 DOI: 10.1016/j.mcpdig.2025.100282

Jeffrey R. Fetzer PhD , Saghir A. Al-Fasly PhD , Cadman L. Leggett MD , Nayantara Coelho-Prabhu MD , Shounak Majumder MD , John B. League III MMIS , Shradha Shalini MS , Ghazal Alabtah , Christine V. Dvorak MAOL , Hamid R. Tizhoosh PhD

引用次数: 0

Feature Selection and Machine Learning Strategies Optimize a Minimal Molecular Assay for Cholangiocarcinoma Subtype 特征选择和机器学习策略优化了胆管癌亚型的最小分子检测

Mayo Clinic Proceedings. Digital health

Pub Date : 2025-12-01 Epub Date: 2025-12-16 DOI: 10.1016/j.mcpdig.2025.100283

Ellen L. Larson MD , Erik Jessen PhD , Dong-Gi Mun PhD , Jennifer L. Tomlinson MD , Amro M. Abdelrahman MBBS, MS , Danielle M. Carlson , Hojjat Salehinejad PhD , Rory L. Smoot MD

引用次数: 0

首页上一页

下一页尾页

类型

全部化学•材料生命科学医学物理工程技术环境•农林材料科学地球科学法学管理学化学环境科学与生态学计算机科学教育学经济学农林科学人文科学生物学数学物理与天体物理心理学综合性期刊其他工业工程理学历史学农学文学信息工程

数据库

全部 ACS Publications Elsevier ieeexplore Springer The Royal Society of Chemistry Wiley

期刊

Mayo Clinic Proceedings. Digital health

全部 Acc. Chem. Res. ACS Applied Bio Materials ACS Appl. Electron. Mater. ACS Appl. Energy Mater. ACS Appl. Mater. Interfaces ACS Appl. Nano Mater. ACS Appl. Polym. Mater. ACS BIOMATER-SCI ENG ACS Catal. ACS Cent. Sci. ACS Chem. Biol. ACS Chemical Health & Safety ACS Chem. Neurosci. ACS Comb. Sci. ACS Earth Space Chem. ACS Energy Lett. ACS Infect. Dis. ACS Macro Lett. ACS Mater. Lett. ACS Med. Chem. Lett. ACS Nano ACS Omega ACS Photonics ACS Sens. ACS Sustainable Chem. Eng. ACS Synth. Biol. Anal. Chem. BIOCHEMISTRY-US Bioconjugate Chem. BIOMACROMOLECULES Chem. Res. Toxicol. Chem. Rev. Chem. Mater. CRYST GROWTH DES ENERG FUEL Environ. Sci. Technol. Environ. Sci. Technol. Lett. Eur. J. Inorg. Chem. IND ENG CHEM RES Inorg. Chem. J. Agric. Food. Chem. J. Chem. Eng. Data J. Chem. Educ. J. Chem. Inf. Model. J. Chem. Theory Comput. J. Med. Chem. J. Nat. Prod. J PROTEOME RES J. Am. Chem. Soc. LANGMUIR MACROMOLECULES Mol. Pharmaceutics Nano Lett. Org. Lett. ORG PROCESS RES DEV ORGANOMETALLICS J. Org. Chem. J. Phys. Chem. J. Phys. Chem. A J. Phys. Chem. B J. Phys. Chem. C J. Phys. Chem. Lett. Analyst Anal. Methods Biomater. Sci. Catal. Sci. Technol. Chem. Commun. Chem. Soc. Rev. CHEM EDUC RES PRACT CRYSTENGCOMM Dalton Trans. Energy Environ. Sci. ENVIRON SCI-NANO ENVIRON SCI-PROC IMP ENVIRON SCI-WAT RES Faraday Discuss. Food Funct. Green Chem. Inorg. Chem. Front. Integr. Biol. J. Anal. At. Spectrom. J. Mater. Chem. A J. Mater. Chem. B J. Mater. Chem. C Lab Chip Mater. Chem. Front. Mater. Horiz. MEDCHEMCOMM Metallomics Mol. Biosyst. Mol. Syst. Des. Eng. Nanoscale Nanoscale Horiz. Nat. Prod. Rep. New J. Chem. Org. Biomol. Chem. Org. Chem. Front. PHOTOCH PHOTOBIO SCI PCCP Polym. Chem.

﹀