Classification of subtypes including LCNEC in lung cancer biopsy slides using convolutional neural network from scratch

• Published in: Scientific reports Vol. 12; no. 1; p. 1830
• Main Authors: Yang, Jung Wook, Song, Dae Hyun, An, Hyo Jung, Seo, Sat Byul
• Format: Journal Article
• Language: English
• Published: England Nature Publishing Group 03.02.2022; Nature Publishing Group UK; Nature Portfolio
• Subjects: Adenocarcinoma of Lung - classification; Adenocarcinoma of Lung - diagnosis; Adenocarcinoma of Lung - pathology; Area Under Curve; Biopsy; Cancer; Carcinoma, Non-Small-Cell Lung - classification; Carcinoma, Non-Small-Cell Lung - diagnosis; Carcinoma, Non-Small-Cell Lung - pathology; Carcinoma, Squamous Cell - classification; Carcinoma, Squamous Cell - diagnosis; Carcinoma, Squamous Cell - pathology; Datasets as Topic; Eosine Yellowish-(YS); Hematoxylin; Histocytochemistry - statistics & numerical data; Humans; Image Interpretation, Computer-Assisted - statistics & numerical data; Image processing; Lung - pathology; Lung cancer; Lung carcinoma; Lung Neoplasms - classification; Lung Neoplasms - diagnosis; Lung Neoplasms - pathology; Neural networks; Neural Networks, Computer; Non-small cell lung carcinoma; Pathology; Small Cell Lung Carcinoma - classification; Small Cell Lung Carcinoma - diagnosis; Small Cell Lung Carcinoma - pathology; Transfer learning; Tumors
• ISSN: 2045-2322; 2045-2322
• DOI: 10.1038/s41598-022-05709-7

Identifying the lung carcinoma subtype in small biopsy specimens is an important part of determining a suitable treatment plan but is often challenging without the help of special and/or immunohistochemical stains. Pathology image analysis that tackles this issue would be helpful for diagnoses and subtyping of lung carcinoma. In this study, we developed AI models to classify multinomial patterns of lung carcinoma; ADC, LCNEC, SCC, SCLC, and non-neoplastic lung tissue based on convolutional neural networks (CNN or ConvNet). Four CNNs that were pre-trained using transfer learning and one CNN built from scratch were used to classify patch images from pathology whole-slide images (WSIs). We first evaluated the diagnostic performance of each model in the test sets. The Xception model and the CNN built from scratch both achieved the highest performance with a macro average AUC of 0.90. The CNN built from scratch model obtained a macro average AUC of 0.97 on the dataset of four classes excluding LCNEC, and 0.95 on the dataset of three subtypes of lung carcinomas; NSCLC, SCLC, and non-tumor, respectively. Of particular note is that the relatively simple CNN built from scratch may be an approach for pathological image analysis.