An Optimal Deep Learning Approach to BCa Tissue Detection using Case Studies

ECTI Transactions on Computer and Information Technology

article

Kasikrit Damkliang¹  Thakerng Wongsirichot¹  Chawisa Khongrak¹  Piyathida Suwannarat¹ 

[1] Prince of Songkla University

关键词: breast cancer tissue;    deep convolutional neural network;    invasive ductal carcinoma;    classification;    cross-validation;   

DOI  :  10.37936/ecti-cit.2023171.250441

学科分类：医学（综合）

来源: Electrical Engineering/Electronics, Computer, Communications and Information Technology Association

【 摘 要 】

In this paper, we present a Deep Learning (DL) model with optimized performance for breast cancer (BCa) tissue classication. A simple DL approach is applied to the analysis of invasive ductal carcinoma (IDC) tissue, which is the most common BCa subtype. Binary classification of non-IDC and IDC tissues is proposed using Convolutional Neural Networks (CNN) in the training and prediction phases. Our trained model achieved F1 and sensitivity scores of 0.88, as well as micro-average values of 0.94 for the area under the curve (AUC) of the receiver operating characteristic (ROC) curve and 0.95 for the area under the precision-recall curve. Since the le size of our model is small, it has the potential for application in real-world scenarios.

【 授权许可】

CC BY-NC-ND