[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-117806-en":3,"doc-seo-117806-105":29,"detail-sidebar-cat-0-en-105":90},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":11,"language":21,"language_code":22,"site_id":23,"html_lang":22,"table_of_contents":24,"faqs":25,"seo_title":26,"seo_description":14,"update_tm":27,"read_time":28},117806,549758146520,"Patrick","https://ap-avatar.wpscdn.com/avatar/80002397d8c0411e94?_k=1775819394049821470",7,"Healthcare","Deploying clinical machine learning? Consider the following","Despite heavy attention and investment in clinical machine learning research, few applications reach large-scale real-world deployment in clinical settings. Research progress is essential, but translation practices are equally critical to enable measurable healthcare impact. The gap between expectation and reality is linked to insufficient consideration of key deployment factors. This work surveys clinical ML practitioners with commercial development experience, identifies core categories of deployment challenges, and uses their insights to support better design and development of clinical ML applications.","Deploying clinical machine learning? Consider the following...*  \nCharles Lu, 1, 3 Ken Chang,2, 3 Praveer Singh,2 Stuart Pomerantz, 1 Sean Doyle, 1 Sujay Kakarmath, 1 Christopher Bridge, 1, 2 Jayashree Kalpathy-Cramer 1, 2  \n1 Massachusetts General Hospital, Boston, MA  \n2 Harvard Medical School, Boston, MA  \n3 Massachusetts Institute of Technology, Cambridge, MA  \narXiv :2 109 .069 19v 3 [ cs .LG] 8 Jun 2023  \nAbstract  \nDespite the intense attention and considerable investment into clinical machine learning research, relatively few applications have been deployed at a large scale in a real-world clinical environment. While research is essential in advancing the stateof-the-art, translation is equally important in bringing these techniques and technologies into a position to impact healthcare ultimately. We believe a lack of appreciation for several considerations is a signi􀀂cant cause for this discrepancy between expectation and reality. To better characterize a holistic perspective among researchers and practitioners, we survey several practitioners with commercial experience developing machine learning tools for clinical deployment. Using these insights, we identify several main categories of challenges to design better and develop clinical machine learning applications.  \nIntroduction  \nAfter the COVID-19 pandemic, hundreds of AI papers were published, but few were clinically practical or useful (Born et al. 2021; Roberts et al. 2021) . Even large technology companies have recently encountered unexpected dif􀀂culties in deploying state-of-the-art research into actual clinical products for healthcare applications (Vincent 2021; Strickland 2019) . We believe there is a disconnect between the clinical machine learning (CML) research community and delivered clinical impact from the translation of useful CML tools (Wolff et al. 2021) . While there have been extensive claims of AI performance being equivalent or superior to human doctors within the research literature (Liu et al. 2019), emerging clinical studies reveal contradictions in those claims of purported super-human performances (Freeman et al. 2021; Zech et al. 2018; Voter et al. 2021) .  \nMachine learning (ML) in the healthcare domain faces numerous challenges, such as immense dif􀀂culty in acquiring and annotating large amounts of medical data, recruiting the necessary clinical expertise in validating models, and integration into existing clinical work􀀃ows and infrastructure (Kelly et al. 2019; Stead 2018) . Developing CML software is considered incredibly challenging compared to traditional software systems (Yang et al. 2020; Lu et al. 2020b;  \n*The authors thank John Chen, James Hillis, and Bernardo Bizzo for their support and helpful discussions.  \n\n| Role | Years of experience | Considerations |\n| --- | --- | --- |\n| Machine learning scientist | 5 | 2.1, 3.2, 4.1, 4.2 |\n| Machine learning scientist | 10 | 1.2, 2.2, 3.1, 3.3 |\n| Software engineer | 30 | 2.4, 3.2, 3.3, 4.1 |\n| Clinical project manager | 5 | 1.1, 1.3, 2.1, 4.2 |\n| Radiologist | 25 | 1.3, 2.3, 4.3 |\n| Neurologist | 10 | 1.3, 2.3 |\n| Clinical researcher | 5 | 2.1, 4.2 |\n\nTable 1: Role and (approximate) experience level of surveyed participants and their (general) contributions to speci􀀂c considerations.  \nHe et al. 2019) . Furthermore, sub-􀀂elds of healthcare such as oncology and dentistry have even more nuanced challenges to adopting AI to their respective specialties (Bi et al. 2019; Schwendicke, Samek, and Krois 2020) . Others have attempted to operationalize these challenges into a formal speci􀀂cation; Oala et al. (2020) apply the ITU/WHO FGAI4H framework to audit several case studies of AI applications for diagnostic retinopathy, Alzheimer’s diagnosis, and cytomorphologic classi􀀂cation for leukemia.  \nTo gain a more holistic understanding of the challenges in CML translation, we survey researchers and clinicians and distill their insights into better practices. The contributions of our survey compl","cbCairJ9xmyGSkVT","https://ap.wps.com/l/cbCairJ9xmyGSkVT","pdf",142486,1,"English","en",105,"# Introduction\n## Clinical research vs real-world impact\n# Considerations\n## Survey methodology\n## Clinical context\n## Clinical validation\n## Deployment\n## Monitoring","[{\"question\":\"Why are relatively few clinical machine learning applications deployed at large scale?\",\"answer\":\"Despite major research attention and investment, deployment in real clinical environments remains limited. The text attributes this gap to inadequate appreciation of key considerations required for translation into practice.\"},{\"question\":\"What approach does the paper use to understand deployment challenges?\",\"answer\":\"It surveys practitioners with commercial experience developing machine learning tools for clinical deployment. Insights from these interviews are synthesized into explicit categories of challenges and better practices.\"},{\"question\":\"What four broad areas do the identified considerations cover?\",\"answer\":\"The considerations are grouped into four areas: clinical context, clinical validation, deployment, and monitoring.\"}]","Deploying clinical machine learning? Consider the following | PDF",1785679668,18,{"code":4,"msg":30,"data":31},"ok",{"site_id":23,"language":22,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":85,"head_meta":87,"extra_data":89,"updated_unix":27},"deploying-clinical-machine-learning-consider-the-following","",{"@graph":35,"@context":84},[36,53,67],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/healthcare/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/deploying-clinical-machine-learning-consider-the-following/117806/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":22,"description":14,"dateModified":61,"datePublished":61,"encodingFormat":60,"isAccessibleForFree":62,"interactionStatistic":63},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-08-02",true,{"@type":64,"interactionType":65,"userInteractionCount":4},"InteractionCounter",{"@type":66},"ViewAction",{"@type":68,"mainEntity":69},"FAQPage",[70,76,80],{"name":71,"@type":72,"acceptedAnswer":73},"Why are relatively few clinical machine learning applications deployed at large scale?","Question",{"text":74,"@type":75},"Despite major research attention and investment, deployment in real clinical environments remains limited. The text attributes this gap to inadequate appreciation of key considerations required for translation into practice.","Answer",{"name":77,"@type":72,"acceptedAnswer":78},"What approach does the paper use to understand deployment challenges?",{"text":79,"@type":75},"It surveys practitioners with commercial experience developing machine learning tools for clinical deployment. Insights from these interviews are synthesized into explicit categories of challenges and better practices.",{"name":81,"@type":72,"acceptedAnswer":82},"What four broad areas do the identified considerations cover?",{"text":83,"@type":75},"The considerations are grouped into four areas: clinical context, clinical validation, deployment, and monitoring.","https://schema.org",{"og:url":51,"og:type":86,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":88,"canonical":51},"index,follow",{"doc_id":7,"site_id":23},{"code":4,"msg":5,"data":91},[92,96,100,104,109,114,117,122,127,130,134],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":93,"show_sort_weight":94,"slug":95},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":97,"show_sort_weight":98,"slug":99},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":101,"show_sort_weight":102,"slug":103},"Exam",70,"exam",{"id":105,"doc_module":4,"doc_module_name":45,"category_name":106,"show_sort_weight":107,"slug":108},5,"Comic",60,"comic",{"id":110,"doc_module":4,"doc_module_name":45,"category_name":111,"show_sort_weight":112,"slug":113},6,"Technology",50,"technology",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":115,"slug":116},40,"healthcare",{"id":118,"doc_module":4,"doc_module_name":45,"category_name":119,"show_sort_weight":120,"slug":121},8,"Research & Report",30,"research-report",{"id":123,"doc_module":4,"doc_module_name":45,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":45,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":45,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":45,"category_name":136,"show_sort_weight":105,"slug":137},19,"General","general"]