Convolutional neural networks can accurately distinguish four histologic growth patterns of lung adenocarcinoma in digital slides

Arkadiusz Gertych,Nathan Ing,Zhaoxuan Ma,Tomasz Markiewicz,Zaneta Swiderska-Chadaj,Szczepan Cierniak,Samuel Guzman,Ann E Walts,Beatrice S Knudsen,Hootan Salemi

doi:10.1038/s41598-018-37638-9

Abstract

During the diagnostic workup of lung adenocarcinomas (LAC), pathologists evaluate distinct histological tumor growth patterns. The percentage of each pattern on multiple slides bears prognostic significance. To assist with the quantification of growth patterns, we constructed a pipeline equipped with a convolutional neural network (CNN) and soft-voting as the decision function to recognize solid, micropapillary, acinar, and cribriform growth patterns, and non-tumor areas. Slides of primary LAC were obtained from Cedars-Sinai Medical Center (CSMC), the Military Institute of Medicine in Warsaw and the TCGA portal. Several CNN models trained with 19,924 image tiles extracted from 78 slides (MIMW and CSMC) were evaluated on 128 test slides from the three sites by F1-score and accuracy using manual tumor annotations by pathologist. The best CNN yielded F1-scores of 0.91 (solid), 0.76 (micropapillary), 0.74 (acinar), 0.6 (cribriform), and 0.96 (non-tumor) respectively. The overall accuracy of distinguishing the five tissue classes was 89.24%. Slide-based accuracy in the CSMC set (88.5%) was significantly better (p < 2.3E-4) than the accuracy in the MIMW (84.2%) and TCGA (84%) sets due to superior slide quality. Our model can work side-by-side with a pathologist to accurately quantify the percentages of growth patterns in tumors with mixed LAC patterns.

Highlights

Lung cancer is currently the second most common cancer in men and women and the leading cause of cancer-related deaths worldwide
Visual assessment of tumor maps outputted by our classification pipeline was performed prior to quantitative evaluation of tumor growth pattern classification results
The tumor maps were false colored and displayed side-by-side to original Hematoxylin and eosin (H&E) slides with superimposed pathologist annotations (Fig. 5)

Summary

Introduction

Lung cancer is currently the second most common cancer in men and women and the leading cause of cancer-related deaths worldwide. Within invasive lung adenocarcinomas (LAC), the new WHO classification separates six histological patterns: lepidic, papillary, micropapillary, acinar, cribriform and solid and recommends that surgically excised tumors be subclassified based on the predominant growth pattern[1]. Based on this recommendation, the histologic patterns observed in the tumor are quantified in 5% increments and reported. Over 80% of LACs demonstrate a mixture of two or more histologic growth patterns, and the evaluation of tumor histology requires a composite manual estimation of the percentage of each pattern in each of several slides prepared from the tumor. CNNs can be instrumental to systematically analyze lung tumors whose histomorphologic heterogeneity poses a challenge to direct visual microscopic quantification of growth patterns by pathologists

Objectives

Methods

Results

Conclusion