Utilizing Automated Breast Cancer Detection to Identify Spatial Distributions of Tumor Infiltrating Lymphocytes in Invasive Breast Cancer

  • 2020-01-13 16:32:45
  • Han Le, Rajarsi Gupta, Le Hou, Shahira Abousamra, Danielle Fassler, Tahsin Kurc, Dimitris Samaras, Rebecca Batiste, Tianhao Zhao, Arvind Rao, Alison L. Van Dyke, Ashish Sharma, Erich Bremer, Jonas S. Almeida, Joel Saltz
  • 0

Abstract

Quantitative assessment of Tumor-TIL spatial relationships is increasinglyimportant in both basic science and clinical aspects of breast cancer research.We have developed and evaluated convolutional neural network (CNN) analysispipelines to generate combined maps of cancer regions and tumor infiltratinglymphocytes (TILs) in routine diagnostic breast cancer whole slide tissueimages (WSIs). We produce interactive whole slide maps that provide 1) insightabout the structural patterns and spatial distribution of lymphocyticinfiltrates and 2) facilitate improved quantification of TILs. We evaluatedboth tumor and TIL analyses using three CNN networks - Resnet-34, VGG16 andInception v4, and demonstrated that the results compared favorably to thoseobtained by what believe are the best published methods. We have producedopen-source tools and generated a public dataset consisting of tumor/TIL mapsfor 1,015 TCGA breast cancer images. We also present a customized web-basedinterface that enables easy visualization and interactive exploration ofhigh-resolution combined Tumor-TIL maps for 1,015TCGA invasive breast cancercases that can be downloaded for further downstream analyses.

 

Quick Read (beta)

This feature is not avaialbe for this paper.