{"id":438,"date":"2025-06-25T17:20:54","date_gmt":"2025-06-25T17:20:54","guid":{"rendered":"https:\/\/d3.harvard.edu\/lish\/?page_id=438"},"modified":"2025-06-25T17:20:54","modified_gmt":"2025-06-25T17:20:54","slug":"driver-of-medical-imaging-diagnoses","status":"publish","type":"page","link":"https:\/\/d3.harvard.edu\/lish\/driver-of-medical-imaging-diagnoses\/","title":{"rendered":"Driver of medical Imaging diagnoses"},"content":{"rendered":"\n<p>LISH works to better understand the variability of doctors\u2019 diagnoses in healthcare, the drivers behind such variability, the impact it has on developing and benchmarking AI systems for healthcare, as well as the methodologies for attaining high quality datasets\u00a0for training AI algorithms to achieve superhuman performance. In addition, we apply such techniques in combination with crowdsourcing to develop AI solutions to medical imaging problems. We also seek to better understand how human-AI interactions impact decision making in healthcare.<\/p>\n\n\n\n<h1 class=\"wp-block-heading\" id=\"page-title\">Dental Image Recognition System&nbsp;<\/h1>\n\n\n\n<p>In collaboration with Charite-Berlin Hospital, we are studying the drivers of variability in doctor performance when diagnosing ailments in dental x-ray images, and how multiple human-labelings of the same data can yield more reliable diagnoses of ailments. These studies aim to provide new insights on improving clinical care and&nbsp;a better understanding biases and shortcomings in evaluating diagnostic imaging. Further, we want to understand if multiple-expert labelling can be used a proxy for clinical ground-truth which should have implications of how to test AI algorithms in medical imaging in general.<\/p>\n\n\n\n<p><\/p>\n\n\n\n<h1 class=\"wp-block-heading\" id=\"page-title\">Lung Cancer Tumor Characterization Image Recognition Algorithm &amp; System&nbsp;<\/h1>\n\n\n\n<p>Lung cancer is the leading cause of cancer death in the United States. Successful treatment depends on a radiation oncologist\u2019s ability to accurately measure the tumor\u2019s shape and responsiveness to interventions. Furthermore, manual delineation of&nbsp;tumors is time consuming and prone to inconsistency or bias. The goal of this project was to produce, through a series of competitions on Topcoder in 2016 and 2017 (see the&nbsp;<a href=\"https:\/\/community.topcoder.com\/longcontest\/?module=ViewProblemStatement&amp;rd=16870&amp;pm=14511\">problem statement for Part 1;<\/a>&nbsp;<a href=\"https:\/\/community.topcoder.com\/longcontest\/?module=ViewProblemStatement&amp;rd=16894&amp;pm=14553\">problem statement for Part 2<\/a>), an automatic tumor delineation algorithm that parallels the accuracy of an average expert while exceeding an expert in terms of speed and consistency.<\/p>\n\n\n\n<p>In Part 1, contestants were tasked with locating and contouring the tumor on images. Thirty-one&nbsp;competitors submitted 244 solutions and competed for $35,000 in prizes. The top solution was able to locate about 70% of the tumors. In Part 2, contestants were asked to improve the contouring performance of their solutions, given a point within the tumor as additional data. Eleven&nbsp;competitors submitted 164 solutions and competed for $15,000. Part 2 yielded a 40% improvement in performance from Part 1. For more information, see&nbsp;<a href=\"https:\/\/www.topcoder.com\/blog\/improve-lung-cancer-treatment-with-algorithms\/\">here<\/a>.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>LISH works to better understand the variability of doctors\u2019 diagnoses in healthcare, the drivers behind such variability, the impact it has on developing and benchmarking AI systems for healthcare, as well as the methodologies for attaining high quality datasets\u00a0for training AI algorithms to achieve superhuman performance. In addition, we apply such techniques in combination with [&hellip;]<\/p>\n","protected":false},"author":19450,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"_trash_the_other_posts":false,"rkv_hide_featured_image":false,"rkv_hide_page_title":false,"rkv_hide_abstract_shapes":true,"rkv_reduce_header":false,"hbsd3_featured_media_url":"","hbsd3_featured_media_url_autoplay":false,"editor_notices":[],"footnotes":""},"rkv-people-shadow":[],"class_list":["post-438","page","type-page","status-publish","hentry"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v27.3 - https:\/\/yoast.com\/product\/yoast-seo-wordpress\/ -->\n<title>Driver of medical Imaging diagnoses - Laboratory for Innovation Science at Harvard<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/d3.harvard.edu\/lish\/driver-of-medical-imaging-diagnoses\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Driver of medical Imaging diagnoses - Laboratory for Innovation Science at Harvard\" \/>\n<meta property=\"og:description\" content=\"LISH works to better understand the variability of doctors\u2019 diagnoses in healthcare, the drivers behind such variability, the impact it has on developing and benchmarking AI systems for healthcare, as well as the methodologies for attaining high quality datasets\u00a0for training AI algorithms to achieve superhuman performance. In addition, we apply such techniques in combination with [&hellip;]\" \/>\n<meta property=\"og:url\" content=\"https:\/\/d3.harvard.edu\/lish\/driver-of-medical-imaging-diagnoses\/\" \/>\n<meta property=\"og:site_name\" content=\"Laboratory for Innovation Science at Harvard\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data1\" content=\"2 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/d3.harvard.edu\\\/lish\\\/driver-of-medical-imaging-diagnoses\\\/\",\"url\":\"https:\\\/\\\/d3.harvard.edu\\\/lish\\\/driver-of-medical-imaging-diagnoses\\\/\",\"name\":\"Driver of medical Imaging diagnoses - Laboratory for Innovation Science at Harvard\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/d3.harvard.edu\\\/lish\\\/#website\"},\"datePublished\":\"2025-06-25T17:20:54+00:00\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/d3.harvard.edu\\\/lish\\\/driver-of-medical-imaging-diagnoses\\\/#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/d3.harvard.edu\\\/lish\\\/driver-of-medical-imaging-diagnoses\\\/\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/d3.harvard.edu\\\/lish\\\/driver-of-medical-imaging-diagnoses\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/d3.harvard.edu\\\/lish\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Driver of medical Imaging diagnoses\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/d3.harvard.edu\\\/lish\\\/#website\",\"url\":\"https:\\\/\\\/d3.harvard.edu\\\/lish\\\/\",\"name\":\"Laboratory for Innovation Science at Harvard\",\"description\":\"\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/d3.harvard.edu\\\/lish\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Driver of medical Imaging diagnoses - Laboratory for Innovation Science at Harvard","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/d3.harvard.edu\/lish\/driver-of-medical-imaging-diagnoses\/","og_locale":"en_US","og_type":"article","og_title":"Driver of medical Imaging diagnoses - Laboratory for Innovation Science at Harvard","og_description":"LISH works to better understand the variability of doctors\u2019 diagnoses in healthcare, the drivers behind such variability, the impact it has on developing and benchmarking AI systems for healthcare, as well as the methodologies for attaining high quality datasets\u00a0for training AI algorithms to achieve superhuman performance. In addition, we apply such techniques in combination with [&hellip;]","og_url":"https:\/\/d3.harvard.edu\/lish\/driver-of-medical-imaging-diagnoses\/","og_site_name":"Laboratory for Innovation Science at Harvard","twitter_card":"summary_large_image","twitter_misc":{"Est. reading time":"2 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/d3.harvard.edu\/lish\/driver-of-medical-imaging-diagnoses\/","url":"https:\/\/d3.harvard.edu\/lish\/driver-of-medical-imaging-diagnoses\/","name":"Driver of medical Imaging diagnoses - Laboratory for Innovation Science at Harvard","isPartOf":{"@id":"https:\/\/d3.harvard.edu\/lish\/#website"},"datePublished":"2025-06-25T17:20:54+00:00","breadcrumb":{"@id":"https:\/\/d3.harvard.edu\/lish\/driver-of-medical-imaging-diagnoses\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/d3.harvard.edu\/lish\/driver-of-medical-imaging-diagnoses\/"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/d3.harvard.edu\/lish\/driver-of-medical-imaging-diagnoses\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/d3.harvard.edu\/lish\/"},{"@type":"ListItem","position":2,"name":"Driver of medical Imaging diagnoses"}]},{"@type":"WebSite","@id":"https:\/\/d3.harvard.edu\/lish\/#website","url":"https:\/\/d3.harvard.edu\/lish\/","name":"Laboratory for Innovation Science at Harvard","description":"","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/d3.harvard.edu\/lish\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"}]}},"_links":{"self":[{"href":"https:\/\/d3.harvard.edu\/lish\/wp-json\/wp\/v2\/pages\/438","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/d3.harvard.edu\/lish\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/d3.harvard.edu\/lish\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/d3.harvard.edu\/lish\/wp-json\/wp\/v2\/users\/19450"}],"replies":[{"embeddable":true,"href":"https:\/\/d3.harvard.edu\/lish\/wp-json\/wp\/v2\/comments?post=438"}],"version-history":[{"count":1,"href":"https:\/\/d3.harvard.edu\/lish\/wp-json\/wp\/v2\/pages\/438\/revisions"}],"predecessor-version":[{"id":439,"href":"https:\/\/d3.harvard.edu\/lish\/wp-json\/wp\/v2\/pages\/438\/revisions\/439"}],"wp:attachment":[{"href":"https:\/\/d3.harvard.edu\/lish\/wp-json\/wp\/v2\/media?parent=438"}],"wp:term":[{"taxonomy":"rkv-people-shadow","embeddable":true,"href":"https:\/\/d3.harvard.edu\/lish\/wp-json\/wp\/v2\/rkv-people-shadow?post=438"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}