Weiyi Wu, Xiaoying Liu, Robert B Hamilton, Arief A Suriawinata, Saeed Hassanpour
{"title":"Graph Convolutional Neural Networks for Histologic Classification of Pancreatic Cancer.","authors":"Weiyi Wu, Xiaoying Liu, Robert B Hamilton, Arief A Suriawinata, Saeed Hassanpour","doi":"10.5858/arpa.2022-0035-OA","DOIUrl":null,"url":null,"abstract":"<p><strong>Context.—: </strong>Pancreatic ductal adenocarcinoma has some of the worst prognostic outcomes among various cancer types. Detection of histologic patterns of pancreatic tumors is essential to predict prognosis and decide the treatment for patients. This histologic classification can have a large degree of variability even among expert pathologists.</p><p><strong>Objective.—: </strong>To detect aggressive adenocarcinoma and less aggressive pancreatic tumors from nonneoplasm cases using a graph convolutional network-based deep learning model.</p><p><strong>Design.—: </strong>Our model uses a convolutional neural network to extract detailed information from every small region in a whole slide image. Then, we use a graph architecture to aggregate the extracted features from these regions and their positional information to capture the whole slide-level structure and make the final prediction.</p><p><strong>Results.—: </strong>We evaluated our model on an independent test set and achieved an F1 score of 0.85 for detecting neoplastic cells and ductal adenocarcinoma, significantly outperforming other baseline methods.</p><p><strong>Conclusions.—: </strong>If validated in prospective studies, this approach has a great potential to assist pathologists in identifying adenocarcinoma and other types of pancreatic tumors in clinical settings.</p>","PeriodicalId":8305,"journal":{"name":"Archives of pathology & laboratory medicine","volume":null,"pages":null},"PeriodicalIF":3.7000,"publicationDate":"2023-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10356903/pdf/","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Archives of pathology & laboratory medicine","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.5858/arpa.2022-0035-OA","RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"MEDICAL LABORATORY TECHNOLOGY","Score":null,"Total":0}
引用次数: 0
Abstract
Context.—: Pancreatic ductal adenocarcinoma has some of the worst prognostic outcomes among various cancer types. Detection of histologic patterns of pancreatic tumors is essential to predict prognosis and decide the treatment for patients. This histologic classification can have a large degree of variability even among expert pathologists.
Objective.—: To detect aggressive adenocarcinoma and less aggressive pancreatic tumors from nonneoplasm cases using a graph convolutional network-based deep learning model.
Design.—: Our model uses a convolutional neural network to extract detailed information from every small region in a whole slide image. Then, we use a graph architecture to aggregate the extracted features from these regions and their positional information to capture the whole slide-level structure and make the final prediction.
Results.—: We evaluated our model on an independent test set and achieved an F1 score of 0.85 for detecting neoplastic cells and ductal adenocarcinoma, significantly outperforming other baseline methods.
Conclusions.—: If validated in prospective studies, this approach has a great potential to assist pathologists in identifying adenocarcinoma and other types of pancreatic tumors in clinical settings.
期刊介绍:
Welcome to the website of the Archives of Pathology & Laboratory Medicine (APLM). This monthly, peer-reviewed journal of the College of American Pathologists offers global reach and highest measured readership among pathology journals.
Published since 1926, ARCHIVES was voted in 2009 the only pathology journal among the top 100 most influential journals of the past 100 years by the BioMedical and Life Sciences Division of the Special Libraries Association. Online access to the full-text and PDF files of APLM articles is free.