T. J. Hart, Chloe Engler Hart, Spencer Hopson, Paul M. Urie, Dennis Della Corte
{"title":"通过更好的数据集和贝叶斯方法克服基于人工智能的前列腺癌检测局限性","authors":"T. J. Hart, Chloe Engler Hart, Spencer Hopson, Paul M. Urie, Dennis Della Corte","doi":"arxiv-2406.06801","DOIUrl":null,"url":null,"abstract":"Despite considerable progress in developing artificial intelligence (AI)\nalgorithms for prostate cancer detection from whole slide images, the clinical\napplicability of these models remains limited due to variability in\npathological annotations and existing dataset limitations. This article\nproposes a novel approach to overcome these challenges by leveraging a Bayesian\nframework to seamlessly integrate new data, and present results as a panel of\nannotations. The framework is demonstrated by integrating a Bayesian prior with\none trained AI model to generate a distribution of Gleason patterns for each\npixel of an image. It is shown that using this distribution of Gleason patterns\nrather than a ground-truth label can improve model applicability, mitigate\nerrors, and highlight areas of interest for pathologists. Additionally, we\npresent a high-quality, hand-curated dataset of prostate histopathological\nimages annotated at the gland level by trained pre-medical students and\nverified by an expert pathologist. We highlight the potential of this adaptive\nand uncertainty-aware framework for developing clinically deployable AI tools\nthat can support pathologists in accurate prostate cancer grading, improve\ndiagnostic accuracy, and create positive patient outcomes.","PeriodicalId":501572,"journal":{"name":"arXiv - QuanBio - Tissues and Organs","volume":"12 1","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2024-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Overcoming Limitations in Artificial Intelligence-based Prostate Cancer Detection through Better Datasets and a Bayesian Approach to Aggregate Panel Predictions\",\"authors\":\"T. J. Hart, Chloe Engler Hart, Spencer Hopson, Paul M. Urie, Dennis Della Corte\",\"doi\":\"arxiv-2406.06801\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Despite considerable progress in developing artificial intelligence (AI)\\nalgorithms for prostate cancer detection from whole slide images, the clinical\\napplicability of these models remains limited due to variability in\\npathological annotations and existing dataset limitations. This article\\nproposes a novel approach to overcome these challenges by leveraging a Bayesian\\nframework to seamlessly integrate new data, and present results as a panel of\\nannotations. The framework is demonstrated by integrating a Bayesian prior with\\none trained AI model to generate a distribution of Gleason patterns for each\\npixel of an image. It is shown that using this distribution of Gleason patterns\\nrather than a ground-truth label can improve model applicability, mitigate\\nerrors, and highlight areas of interest for pathologists. Additionally, we\\npresent a high-quality, hand-curated dataset of prostate histopathological\\nimages annotated at the gland level by trained pre-medical students and\\nverified by an expert pathologist. We highlight the potential of this adaptive\\nand uncertainty-aware framework for developing clinically deployable AI tools\\nthat can support pathologists in accurate prostate cancer grading, improve\\ndiagnostic accuracy, and create positive patient outcomes.\",\"PeriodicalId\":501572,\"journal\":{\"name\":\"arXiv - QuanBio - Tissues and Organs\",\"volume\":\"12 1\",\"pages\":\"\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-06-10\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - QuanBio - Tissues and Organs\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2406.06801\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - QuanBio - Tissues and Organs","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2406.06801","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Overcoming Limitations in Artificial Intelligence-based Prostate Cancer Detection through Better Datasets and a Bayesian Approach to Aggregate Panel Predictions
Despite considerable progress in developing artificial intelligence (AI)
algorithms for prostate cancer detection from whole slide images, the clinical
applicability of these models remains limited due to variability in
pathological annotations and existing dataset limitations. This article
proposes a novel approach to overcome these challenges by leveraging a Bayesian
framework to seamlessly integrate new data, and present results as a panel of
annotations. The framework is demonstrated by integrating a Bayesian prior with
one trained AI model to generate a distribution of Gleason patterns for each
pixel of an image. It is shown that using this distribution of Gleason patterns
rather than a ground-truth label can improve model applicability, mitigate
errors, and highlight areas of interest for pathologists. Additionally, we
present a high-quality, hand-curated dataset of prostate histopathological
images annotated at the gland level by trained pre-medical students and
verified by an expert pathologist. We highlight the potential of this adaptive
and uncertainty-aware framework for developing clinically deployable AI tools
that can support pathologists in accurate prostate cancer grading, improve
diagnostic accuracy, and create positive patient outcomes.