Deep learning-based six-type classifier for lung cancer and mimics from histopathological whole slide images: a retrospective study

Abstract

Background Targeted therapy and immunotherapy put forward higher demands for accurate lung cancer classification, as well as benign versus malignant disease discrimination. Digital whole slide images (WSIs) witnessed the transition from traditional histopathology to computational approaches, arousing a hype of deep learning methods for histopathological analysis. We aimed at exploring the potential of deep learning models in the identification of lung cancer subtypes and cancer mimics from WSIs. Methods We initially obtained 741 WSIs from the First Affiliated Hospital of Sun Yat-sen University (SYSUFH) for the deep learning model development, optimization, and verification. Additional 318 WSIs from SYSUFH, 212 from Shenzhen People’s Hospital, and 422 from The Cancer Genome Atlas were further collected for multi-centre verification. EfficientNet-B5- and ResNet-50-based deep learning methods were developed and compared using the metrics of recall, precision, F1-score, and areas under the curve (AUCs). A threshold-based tumour-first aggregation approach was proposed and implemented for the label inferencing of WSIs with complex tissue components. Four pathologists of different levels from SYSUFH reviewed all the testing slides blindly, and the diagnosing results were used for quantitative comparisons with the best performing deep learning model. Results We developed the first deep learning-based six-type classifier for histopathological WSI classification of lung adenocarcinoma, lung squamous cell carcinoma, small cell lung carcinoma, pulmonary tuberculosis, organizing pneumonia, and normal lung. The EfficientNet-B5-based model outperformed ResNet-50 and was selected as the backbone in the classifier. Tested on 1067 slides from four cohorts of different medical centres, AUCs of 0.970, 0.918, 0.963, and 0.978 were achieved, respectively. The classifier achieved high consistence to the ground truth and attending pathologists with high intraclass correlation coefficients over 0.873. Conclusions Multi-cohort testing demonstrated our six-type classifier achieved consistent and comparable performance to experienced pathologists and gained advantages over other existing computational methods. The visualization of prediction heatmap improved the model interpretability intuitively. The classifier with the threshold-based tumour-first label inferencing method exhibited excellent accuracy and feasibility in classifying lung cancers and confused nonneoplastic tissues, indicating that deep learning can resolve complex multi-class tissue classification that conforms to real-world histopathological scenarios.

NECScanNet: Novel Method for Cervical Neuroendocrine Cancer Screening from Whole Slide Images

“…Learning. Multiple-instance learning (MIL) is a weakly supervised learning, which establishes a multi-instance classifier by learning word bags with category labels and applies the classifier to the prediction of unknown word bags [32,33].…”

Deep Learning Analysis of CT Images Reveals High-Grade Pathological Features to Predict Survival in Lung Adenocarcinoma

“…There have been many studies on the imaging-based deep learning model for predicting the histology of lung cancer (i.e., adenocarcinoma, squamous cell carcinoma, and small cell lung cancer), and it has shown relatively good accuracy [27][28][29][30]. However, there are few studies on radiomics-based machine learning or CT-based deep learning models predicting high grade histologic patterns of lung ADC (i.e., micropapillary or solid pattern, MPSol).…”

Applications of Artificial Intelligence (AI) in medicine

“…Prowadzonych jest coraz więcej badań, których celem jest porównanie trafności rozpoznań dokonywanych przez sztuczną inteligencję z rozpoznaniami stawianymi przez doświadczonych patomorfologów. Wyniki dowodzą, że w dużej części przypadków sztuczna inteligencja radzi sobie podobnie do specjalistów, a dodatkowo może im pomagać w stawianiu rozpoznań, zmniejszając tym samym ryzyko błędu ludzkiego [43][44][45][46][47][48].…”

“…Oparty na głębokim uczeniu klasyfikator rozróżniający pod względem histopatologicznym gruczolakoraka płuc, raka płaskonabłonkowego płuc, raka drobnokomórkowego płuc, gruźlicę płuc, organizujące się zapalenie płuc i prawidłowe płuca, przetestowany na 1067 preparatach mikroskopowych pochodzących z 4 kohort z różnych ośrodków medycznych, uzyskał AUC wynoszące kolejno: 0,970, 0,918, 0,963, 0,978. Współczynnik zgodności z klasyfikacją przeprowadzoną przez doświadczonych patomorfologów wyniósł 0,873 [46].…”