Mayo Clinic Proceedings. Digital health最新文献

{"title":"Toward Safe and Ethical Implementation of Health Care Artificial Intelligence: Insights From an Academic Medical Center","authors":"Austin M. Stroud MA ,&nbsp;Michele D. Anzabi MBE ,&nbsp;Journey L. Wise BA ,&nbsp;Barbara A. Barry PhD ,&nbsp;Momin M. Malik PhD ,&nbsp;Michelle L. McGowan PhD ,&nbsp;Richard R. Sharp PhD","doi":"10.1016/j.mcpdig.2024.100189","DOIUrl":"10.1016/j.mcpdig.2024.100189","url":null,"abstract":"<div><div>Claims abound that advances in artificial intelligence (AI) will permeate virtually every aspect of medicine and transform clinical practice. Simultaneously, concerns about the safety and equity of health care AI have prompted ethical and regulatory scrutiny from multiple oversight bodies. Positioned at the intersection of these perspectives, academic medical centers (AMCs) are charged with navigating the safe and responsible implementation of health care AI. Decisions about the use of AI at AMCs are complicated by uncertainties regarding the risks posed by these technologies and a lack of consensus on best practices for managing these risks. In this article, we highlight several potential harms that may arise in the adoption of health care AI, with a focus on risks to patients, clinicians, and medical practice. In addition, we describe several strategies that AMCs might adopt now to address concerns about the safety and ethical uses of health care AI. Our analysis aims to support AMCs as they seek to balance AI innovation with proactive oversight.</div></div>","PeriodicalId":74127,"journal":{"name":"Mayo Clinic Proceedings. Digital health","volume":"3 1","pages":"Article 100189"},"PeriodicalIF":0.0,"publicationDate":"2024-12-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143149187","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Implementing Artificial Intelligence Algorithms in the Radiology Workflow: Challenges and Considerations","authors":"Panagiotis Korfiatis PhD ,&nbsp;Timothy L. Kline PhD ,&nbsp;Holly M. Meyer MS ,&nbsp;Sana Khalid MS ,&nbsp;Timothy Leiner MD ,&nbsp;Brenna T. Loufek MS ,&nbsp;Daniel Blezek PhD ,&nbsp;David E. Vidal JD ,&nbsp;Robert P. Hartman MD ,&nbsp;Lori J. Joppa MBA ,&nbsp;Andrew D. Missert PhD ,&nbsp;Theodora A. Potretzke MD ,&nbsp;Jerome P. Taubel ,&nbsp;Jason A. Tjelta BS ,&nbsp;Matthew R. Callstrom MD ,&nbsp;Eric E. Williamson MD","doi":"10.1016/j.mcpdig.2024.100188","DOIUrl":"10.1016/j.mcpdig.2024.100188","url":null,"abstract":"<div><div>Integration of AI-enabled algorithms into the radiology workflow presents a complex array of challenges that span operational, technical, clinical, and regulatory domains. Successfully overcoming these hurdles requires a multifaceted approach, including strategic planning, educational initiatives, and careful consideration of the practical implications for radiologists' workloads. Institutions must navigate these challenges with a clear understanding of the potential benefits and limitations of both vended and in-house developed AI tools.</div></div>","PeriodicalId":74127,"journal":{"name":"Mayo Clinic Proceedings. Digital health","volume":"3 1","pages":"Article 100188"},"PeriodicalIF":0.0,"publicationDate":"2024-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143149188","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Clinical Applications of Fractional Flow Reserve Derived from Computed Tomography in Coronary Artery Disease","authors":"Cappi Chan MSc ,&nbsp;Min Wang PhD ,&nbsp;Luoyi Kong MSc ,&nbsp;Leanne Li MSc ,&nbsp;Lawrence Wing Chi Chan PhD","doi":"10.1016/j.mcpdig.2024.100187","DOIUrl":"10.1016/j.mcpdig.2024.100187","url":null,"abstract":"<div><div>Computer tomography–derived fractional flow reserve (CT-FFR) represents a significant advancement in noninvasive cardiac functional assessment. This technology uses computer simulation and anatomical information from computer tomography of coronary angiogram to calculate the CT-FFR value at each point within the coronary vasculature. These values serve as a critical reference for cardiologists in making informed treatment decisions and planning. Emerging evidence suggests that CT-FFR has the potential to enhance the specificity of computer tomography of coronary angiogram, thereby reducing the need for additional diagnostic examinations such as invasive coronary angiography and cardiac magnetic resonance imaging. This could result in savings in financial cost, time, and resources for both patients and health care providers. However, it is important to note that although CT-FFR holds great promise, there are limitations to this technology. Users should be cautious of common pitfalls associated with its use. A comprehensive understanding of these limitations is essential for effectively applying CT-FFR in clinical practice.</div></div>","PeriodicalId":74127,"journal":{"name":"Mayo Clinic Proceedings. Digital health","volume":"3 1","pages":"Article 100187"},"PeriodicalIF":0.0,"publicationDate":"2024-12-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143149192","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Developing a Research Center for Artificial Intelligence in Medicine","authors":"Curtis P. Langlotz MD, PhD ,&nbsp;Johanna Kim MPH, MBA ,&nbsp;Nigam Shah MBBS, PhD ,&nbsp;Matthew P. Lungren MD, MPH ,&nbsp;David B. Larson MD, MBA ,&nbsp;Somalee Datta PhD ,&nbsp;Fei Fei Li PhD ,&nbsp;Ruth O’Hara PhD ,&nbsp;Thomas J. Montine MD, PhD ,&nbsp;Robert A. Harrington MD ,&nbsp;Garry E. Gold MD, MS","doi":"10.1016/j.mcpdig.2024.07.005","DOIUrl":"10.1016/j.mcpdig.2024.07.005","url":null,"abstract":"<div><div>Artificial intelligence (AI) and machine learning (ML) are driving innovation in biosciences and are already affecting key elements of medical scholarship and clinical care. Many schools of medicine are capitalizing on the promise of these new technologies by establishing academic units to catalyze and grow research and innovation in AI/ML. At Stanford University, we have developed a successful model for an AI/ML research center with support from academic leaders, clinical departments, extramural grants, and industry partners. The Center for Artificial Intelligence in Medicine and Imaging uses the following 4 key tactics to support AI/ML research: project-based learning opportunities that build interdisciplinary collaboration; internal grant programs that catalyze extramural funding; infrastructure that facilitates the rapid creation of large multimodal AI-ready clinical data sets; and educational and open data programs that engage the broader research community. The center is based on the premise that foundational and applied research are not in tension but instead are complementary. Solving important biomedical problems with AI/ML requires high-quality foundational team science that incorporates the knowledge and expertise of clinicians, clinician scientists, computer scientists, and data scientists. As AI/ML becomes an essential component of research and clinical care, multidisciplinary centers of excellence in AI/ML will become a key part of the scholarly portfolio of academic medical centers and will provide a foundation for the responsible, ethical, and fair implementation of AI/ML systems.</div></div>","PeriodicalId":74127,"journal":{"name":"Mayo Clinic Proceedings. Digital health","volume":"2 4","pages":"Pages 677-686"},"PeriodicalIF":0.0,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142744031","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Digital Health Interventions Supporting Recovery for Intensive Care Patients and Their Family Members: A Scoping Review","authors":"Elke Berger MSc ,&nbsp;Carola Schol MSc ,&nbsp;Sabrina Meertens-Gunput PhD ,&nbsp;Dorien Kiers MD, PhD ,&nbsp;Diederik Gommers MD, PhD ,&nbsp;Louise Rose PhD ,&nbsp;Margo van Mol PhD","doi":"10.1016/j.mcpdig.2024.11.006","DOIUrl":"10.1016/j.mcpdig.2024.11.006","url":null,"abstract":"<div><div>Digital innovation in interventions to promote recovery for intensive care unit (ICU) patients and their family members holds promise for enhancing accessibility and improving physical, psychological, and cognitive outcomes. This scoping review provides a comprehensive overview of digital health interventions designed to support the recovery of ICU patients and their family members described in peer-reviewed publications. We searched 6 databases (inception to September 2023); 2 reviewers independently screened citations against predefined eligibility criteria and extracted data. We screened 3485 records and identified 18 original studies and 8 study protocols with a range of study designs published between 2016 and 2023. Most (n=15) completed studies recruited patients only. Digital interventions were delivered through applications, virtual reality, videoconferencing, and smartwatches. In the completed studies, outcomes are described as feasibility, intervention efficacy, or both. Digital interventions supplemented with professional support and personalized feedback were more feasible than self-directed interventions. Further research is essential to ascertain the efficacy and cost-effectiveness of digital interventions in improving outcomes for ICU survivors and their family members.</div></div>","PeriodicalId":74127,"journal":{"name":"Mayo Clinic Proceedings. Digital health","volume":"3 1","pages":"Article 100185"},"PeriodicalIF":0.0,"publicationDate":"2024-11-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143149193","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Fine-Tuning Large Language Models for Specialized Use Cases","authors":"D.M. Anisuzzaman PhD,&nbsp;Jeffrey G. Malins PhD,&nbsp;Paul A. Friedman MD,&nbsp;Zachi I. Attia PhD","doi":"10.1016/j.mcpdig.2024.11.005","DOIUrl":"10.1016/j.mcpdig.2024.11.005","url":null,"abstract":"<div><div>Large language models (LLMs) are a type of artificial intelligence, which operate by predicting and assembling sequences of words that are statistically likely to follow from a given text input. With this basic ability, LLMs are able to answer complex questions and follow extremely complex instructions. Products created using LLMs such as ChatGPT by OpenAI and Claude by Anthropic have created a huge amount of traction and user engagements and revolutionized the way we interact with technology, bringing a new dimension to human-computer interaction. Fine-tuning is a process in which a pretrained model, such as an LLM, is further trained on a custom data set to adapt it for specialized tasks or domains. In this review, we outline some of the major methodologic approaches and techniques that can be used to fine-tune LLMs for specialized use cases and enumerate the general steps required for carrying out LLM fine-tuning. We then illustrate a few of these methodologic approaches by describing several specific use cases of fine-tuning LLMs across medical subspecialties. Finally, we close with a consideration of some of the benefits and limitations associated with fine-tuning LLMs for specialized use cases, with an emphasis on specific concerns in the field of medicine.</div></div>","PeriodicalId":74127,"journal":{"name":"Mayo Clinic Proceedings. Digital health","volume":"3 1","pages":"Article 100184"},"PeriodicalIF":0.0,"publicationDate":"2024-11-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143148329","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Development of a Human Factors–Based Guideline to Support the Design, Evaluation, and Continuous Improvement of Clinical Decision Support","authors":"Selvana Awad BPharm, MHSM ,&nbsp;Thomas Loveday MPsych, PhD ,&nbsp;Richard Lau BPsychSc ,&nbsp;Melissa T. Baysari BPsych, PhD","doi":"10.1016/j.mcpdig.2024.11.003","DOIUrl":"10.1016/j.mcpdig.2024.11.003","url":null,"abstract":"<div><h3>Objective</h3><div>To develop a vendor-agnostic, human factors (HF)-based guideline to guide the design, evaluation, and continuous improvement of clinical decision support (CDS).</div></div><div><h3>Participants and Methods</h3><div>The study used a 2-phased iterative approach between June 2022 and June 2024. Phase 1 involved a search for relevant industry standards and literature and consultation with multidisciplinary subject matter experts. Phase 2 involved a workshop with 30 health care and academic stakeholders to evaluate face validity and perceived usefulness of the initial section of the guideline. Participants were asked if the guideline met their expectations, to report on usefulness and ease of use and to suggest areas for improvement.</div></div><div><h3>Results</h3><div>Phase 1 resulted in a compilation of accessible, best practice, and context-appropriate HF guidance for CDS design and optimization. The guideline supports users in determining whether use of CDS is appropriate, and if yes, CDS options and design guidance. During phase 2, the guideline addressed 15 of participants’ 19 expectations for a CDS guideline. Participants said the guideline was helpful, comprehensive, easy to use, and provided step-by-step guidance, boundaries, and transparency around CDS decisions. Participants recommended strengthening guidance around the need to understand system capabilities and the technical burden or complexity of CDS, and further guidance on how to approach CDS optimization using the guideline.</div></div><div><h3>Conclusion</h3><div>The 2-phased iterative development and feedback process resulted in the development of an HF-informed guideline to provide consolidated, accessible, and current best practice guidance on the appropriateness of CDS and CDS options, as well as designing, evaluating, and continuously improving CDS. Future work will evaluate the impact and implementation of the guideline in real-world settings.</div></div>","PeriodicalId":74127,"journal":{"name":"Mayo Clinic Proceedings. Digital health","volume":"3 1","pages":"Article 100182"},"PeriodicalIF":0.0,"publicationDate":"2024-11-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143149196","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"","authors":"","doi":"10.1016/j.mcpdig.2024.11.007","DOIUrl":"10.1016/j.mcpdig.2024.11.007","url":null,"abstract":"","PeriodicalId":74127,"journal":{"name":"Mayo Clinic Proceedings. Digital health","volume":"3 1","pages":"Article 100186"},"PeriodicalIF":0.0,"publicationDate":"2024-11-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143148331","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Gait Speed and Task Specificity in Predicting Lower-Limb Kinematics: A Deep Learning Approach Using Inertial Sensors","authors":"Vaibhav R. Shah MSc ,&nbsp;Philippe C. Dixon PhD","doi":"10.1016/j.mcpdig.2024.11.004","DOIUrl":"10.1016/j.mcpdig.2024.11.004","url":null,"abstract":"<div><h3>Objective</h3><div>To develop a deep learning framework to predict lower-limb joint kinematics from inertial measurement unit (IMU) data across multiple gait tasks (walking, jogging, and running) and evaluate the impact of dynamic time warping (DTW) on reducing prediction errors.</div></div><div><h3>Patients and Methods</h3><div>Data were collected from 18 participants fitted with IMUs and an optical motion capture system between May 25, 2023, and May 30, 2023. A long short-term memory autoencoder supervised regression model was developed. The model consisted of multiple long short-term memory and convolution layers. Acceleration and gyroscope data from the IMUs in 3 axes and their magnitude for the proximal and distal sensors of each joint (hip, knee, and ankle) were inputs to the model. Optical motion capture kinematics were considered ground truth and used as an output to train the prediction model.</div></div><div><h3>Results</h3><div>The deep learning models achieved a root-mean-square error of less than 6° for hip, knee, and ankle joint sagittal plane angles, with the ankle showing the lowest error (5.1°). Task-specific models reported enhanced performance during certain gait phases, such as knee flexion during running. The application of DTW significantly reduced root-mean-square error across all tasks by at least 3° to 4°. External validation of independent data confirmed the model’s generalizability.</div></div><div><h3>Conclusion</h3><div>Our findings underscore the potential of IMU-based deep learning models for joint kinematic predictions, offering a practical solution for remote and continuous biomechanical assessments in health care and sports science.</div></div>","PeriodicalId":74127,"journal":{"name":"Mayo Clinic Proceedings. Digital health","volume":"3 1","pages":"Article 100183"},"PeriodicalIF":0.0,"publicationDate":"2024-11-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143149197","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The Effectiveness of Telehealth Intervention on Chronic Kidney Disease Management in Adults: A Systematic Review","authors":"Tess Ellis MS, RD,&nbsp;Anna J. Kwon MS,&nbsp;Mee Young Hong PhD","doi":"10.1016/j.mcpdig.2024.11.002","DOIUrl":"10.1016/j.mcpdig.2024.11.002","url":null,"abstract":"<div><h3>Objective</h3><div>To evaluate the effectiveness of telehealth programs on dietary habits, quality of life, renal function, and blood pressure in adults with chronic kidney disease (CKD).</div></div><div><h3>Patients and Methods</h3><div>A systematic literature review was completed according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines. Using PubMed/Medline, Scopus, Embase, and ScienceDirect databases, articles published between 2012 and 2024 were selected using the following keywords: <em>telehealth</em>, <em>eHealth</em>, <em>mHealth</em>, <em>telemedicine</em>, <em>telenutrition</em>, and <em>chronic kidney disease</em>.</div></div><div><h3>Results</h3><div>A total of 13 studies—10 randomized controlled trials and 3 single-arm trials—were chosen for this review. In these trials, telehealth interventions were administered using mobile applications, phone calls, web-based communications, text messaging, wearable devices, or a combination of these tools to provide treatment for adults with CKD. Interdisciplinary collaboration between a dietitian and other health care team members was shown to improve renal function and dietary habits when providing telehealth interventions via mobile applications, phone calls, and text messaging. Web-based telehealth delivery that involves diverse health care personnel has been shown to improve the quality of life in adult patients with CKD.</div></div><div><h3>Conclusion</h3><div>Receiving treatment using telehealth communication methods may be a beneficial option for adult patients with CKD by enhancing accessibility, promoting multidisciplinary collaboration, and effectively managing blood pressure and dietary habits, leading to improved quality of life for patients. Future research administering homogeneous and rigorously controlled experimental methods with larger and more diverse populations, as well as longer study durations, is necessary to further elucidate the effectiveness of CKD treatment delivery via telehealth for adult patients.</div></div>","PeriodicalId":74127,"journal":{"name":"Mayo Clinic Proceedings. Digital health","volume":"3 1","pages":"Article 100181"},"PeriodicalIF":0.0,"publicationDate":"2024-11-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143149195","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}