Large Language Models (LLMs) are widely used in healthcare, but limitations like hallucinations, incomplete information, and bias hinder their reliability. To address these, researchers released the Build Your Own expert Bot (BYOeB) platform, enabling developers to create LLM-powered chatbots with integrated expert verification. CataractBot, its first implementation, provides expert-verified responses to cataract surgery questions. A pilot evaluation showed its potential; however the study had a small sample size and was primarily qualitative. In this work, we conducted a large-scale 24-week deployment of CataractBot involving 318 patients and attendants who sent 1,992 messages, with 91.71% of responses verified by seven experts. Analysis of interaction logs revealed that medical questions significantly outnumbered logistical ones, hallucinations were negligible, and experts rated 84.52% of medical answers as accurate. As the knowledge base expanded with expert corrections, system performance improved by 19.02%, reducing expert workload. These insights guide the design of future LLM-powered chatbots.<\/p>\n","protected":false},"excerpt":{"rendered":"
Large Language Models (LLMs) are widely used in healthcare, but limitations like hallucinations, incomplete information, and bias hinder their reliability. To address these, researchers released the Build Your Own expert Bot (BYOeB) platform, enabling developers to create LLM-powered chatbots with integrated expert verification. CataractBot, its first implementation, provides expert-verified responses to cataract surgery questions. A […]<\/p>\n","protected":false},"featured_media":0,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"_classifai_error":"","msr-author-ordering":null,"msr_publishername":"","msr_publisher_other":"","msr_booktitle":"","msr_chapter":"","msr_edition":"","msr_editors":"","msr_how_published":"arXiv","msr_isbn":"","msr_issue":"","msr_journal":"","msr_number":"","msr_organization":"","msr_pages_string":"","msr_page_range_start":"","msr_page_range_end":"","msr_series":"","msr_volume":"","msr_copyright":"","msr_conference_name":"","msr_doi":"","msr_arxiv_id":"","msr_s2_paper_id":"","msr_mag_id":"","msr_pubmed_id":"","msr_other_authors":"","msr_other_contributors":"","msr_speaker":"","msr_award":"","msr_affiliation":"","msr_institution":"","msr_host":"","msr_version":"","msr_duration":"","msr_original_fields_of_study":null,"msr_release_tracker_id":"","msr_s2_match_type":"","msr_citation_count_updated":"","msr_published_date":"2024-9-15","msr_highlight_text":"","msr_notes":"","msr_longbiography":"","msr_publicationurl":"","msr_external_url":"","msr_secondary_video_url":"","msr_conference_url":"","msr_journal_url":"","msr_s2_pdf_url":"","msr_year":0,"msr_citation_count":0,"msr_influential_citations":0,"msr_reference_count":0,"msr_s2_match_confidence":0,"msr_microsoftintellectualproperty":true,"msr_s2_open_access":false,"msr_s2_author_ids":[],"msr_pub_ids":[],"msr_hide_image_in_river":null,"footnotes":""},"msr-research-highlight":[],"research-area":[13556,13554,13553],"msr-publication-type":[193724],"msr-publisher":[],"msr-focus-area":[],"msr-locale":[268875],"msr-post-option":[],"msr-field-of-study":[252883,246691,248485],"msr-conference":[],"msr-journal":[],"msr-impact-theme":[261673],"msr-pillar":[],"class_list":["post-1085784","msr-research-item","type-msr-research-item","status-publish","hentry","msr-research-area-artificial-intelligence","msr-research-area-human-computer-interaction","msr-research-area-medical-health-genomics","msr-locale-en_us","msr-field-of-study-chatbot","msr-field-of-study-computer-science","msr-field-of-study-human-computer-interaction"],"msr_publishername":"","msr_edition":"","msr_affiliation":"","msr_published_date":"2024-9-15","msr_host":"","msr_duration":"","msr_version":"","msr_speaker":"","msr_other_contributors":"","msr_booktitle":"","msr_pages_string":"","msr_chapter":"","msr_isbn":"","msr_journal":"","msr_volume":"","msr_number":"","msr_editors":"","msr_series":"","msr_issue":"","msr_organization":"","msr_how_published":"arXiv","msr_notes":"","msr_highlight_text":"","msr_release_tracker_id":"","msr_original_fields_of_study":"","msr_download_urls":"","msr_external_url":"","msr_secondary_video_url":"","msr_longbiography":"","msr_microsoftintellectualproperty":1,"msr_main_download":"","msr_publicationurl":"","msr_doi":"","msr_publication_uploader":[{"type":"url","viewUrl":"false","id":"false","title":"https:\/\/arxiv.org\/abs\/2409.10354","label_id":"243109","label":0}],"msr_related_uploader":"","msr_citation_count":0,"msr_citation_count_updated":"","msr_s2_paper_id":"","msr_influential_citations":0,"msr_reference_count":0,"msr_arxiv_id":"","msr_s2_author_ids":[],"msr_s2_open_access":false,"msr_s2_pdf_url":null,"msr_attachments":[],"msr-author-ordering":[{"type":"text","value":"Bhuvan Sachdeva","user_id":0,"rest_url":false},{"type":"guest","value":"pragnya-ramjee","user_id":1080645,"rest_url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=pragnya-ramjee"},{"type":"text","value":"Geeta Fulari","user_id":0,"rest_url":false},{"type":"guest","value":"dr-kaushik-murali","user_id":914346,"rest_url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=dr-kaushik-murali"},{"type":"user_nicename","value":"Mohit Jain","user_id":38769,"rest_url":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/microsoft-research\/v1\/researchers?person=Mohit Jain"}],"msr_impact_theme":["Health"],"msr_research_lab":[199562],"msr_event":[],"msr_group":[],"msr_project":[979329,979335],"publication":[],"video":[],"msr-tool":[],"msr_publication_type":"miscellaneous","related_content":{"projects":[{"ID":979329,"post_title":"Healthcare Research at MSR India","post_name":"healthcare","post_type":"msr-project","post_date":"2023-11-26 21:16:53","post_modified":"2024-09-19 16:08:15","post_status":"publish","permalink":"https:\/\/www.microsoft.com\/en-us\/research\/project\/healthcare\/","post_excerpt":"At Microsoft Research Lab India, we conduct a variety of healthcare related research, including smartphone-based low-cost diagnostics, generative AI chatbots to support the healthcare ecosystem, and promote mental well-being of employees.","_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project\/979329"}]}},{"ID":979335,"post_title":"HealthBots","post_name":"health-bots","post_type":"msr-project","post_date":"2023-11-26 21:00:15","post_modified":"2023-12-18 10:53:47","post_status":"publish","permalink":"https:\/\/www.microsoft.com\/en-us\/research\/project\/health-bots\/","post_excerpt":"Recent studies have highlighted that more than 70% of patients and their caregivers experience anxiety prior to undergoing an invasive treatment. Additionally, over 80% of them require timely, trustworthy, detailed, and accurate information about their treatment. The provision of such information holds the potential to alleviate pre- and post-operative anxiety. To address this issue, we designed and developed chatbots, powered by state-of-the-art generative AI models fine-tuned on the doctor\u2019s provided knowledge base. These HealthBots aims…","_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project\/979335"}]}}]},"_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/1085784","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item"}],"about":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-research-item"}],"version-history":[{"count":1,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/1085784\/revisions"}],"predecessor-version":[{"id":1085787,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-item\/1085784\/revisions\/1085787"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=1085784"}],"wp:term":[{"taxonomy":"msr-research-highlight","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-research-highlight?post=1085784"},{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=1085784"},{"taxonomy":"msr-publication-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-publication-type?post=1085784"},{"taxonomy":"msr-publisher","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-publisher?post=1085784"},{"taxonomy":"msr-focus-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-focus-area?post=1085784"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=1085784"},{"taxonomy":"msr-post-option","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-post-option?post=1085784"},{"taxonomy":"msr-field-of-study","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-field-of-study?post=1085784"},{"taxonomy":"msr-conference","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-conference?post=1085784"},{"taxonomy":"msr-journal","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-journal?post=1085784"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=1085784"},{"taxonomy":"msr-pillar","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-pillar?post=1085784"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}