Clinically applicable histopathological diagnosis system for gastric cancer detection using deep learning

Abstract

The early detection and accurate histopathological diagnosis of gastric cancer increase the chances of successful treatment. The worldwide shortage of pathologists offers a unique opportunity for the use of artificial intelligence assistance systems to alleviate the workload and increase diagnostic accuracy. Here, we report a clinically applicable system developed at the Chinese PLA General Hospital, China, using a deep convolutional neural network trained with 2,123 pixel-level annotated H&E-stained whole slide images. The model achieves a sensitivity near 100% and an average specificity of 80.6% on a real-world test dataset with 3,212 whole slide images digitalized by three scanners. We show that the system could aid pathologists in improving diagnostic accuracy and preventing misdiagnoses. Moreover, we demonstrate that our system performs robustly with 1,582 whole slide images from two other medical centres. Our study suggests the feasibility and benefits of using histopathological artificial intelligence assistance systems in routine practice scenarios.

Status quo and future prospects of artificial neural network from the perspective of gastroenterologists

“…Haematoxylin-eosin (H&E) staining is the most common method for discriminating benign and malignant lesions. For GC, most studies used CNN models to construct diagnostic systems for H&E staining[ 13 - 17 ]. In 2017, Sharma et al [ 17 ] reported the first study about computer-aided classification in H&E staining of GC.…”

“…Rapid progress has been made during the past three years. A CNN model fed by over 2000 H&E slide images achieved a near 100% sensitivity and 80.6% specificity[ 13 ]. Such optimal sensitivity showed that this model has potential in preliminary screening of H&E staining.…”

Role of artificial intelligence in multidisciplinary imaging diagnosis of gastrointestinal diseases

“…A number of recent studies have paid attention to lesion detection, classification and characterization in this anatomic site. For instance, Song et al [ 84 ] developed and trained a deep CNN to differentiate between benign and malignant gastric tumors, with a sensitivity of 100% and a specificity of 80.6%[ 84 ]. Also, a network developed by Sharma et al [ 85 ] classified gastric cancer cases according to immunohistochemical response and presence of necrosis, with accuracy rates of 0.699 and 0.814, respectively[ 85 ].…”

Clinically applicable Gleason grading (GD) system for prostate cancer based on deep learning

“…As shown in Figure 1 , we used the DeepLab v3 image segmentation model with ResNet-50 to establish the GD system. [ 4 ] During the model training process, the parameters of the gastric cancer detection model [ 5 ] were used as the initial values, and the model parameters were fine-tuned using the prostate training data by transfer learning. The model training was performed with TensorFlow on 8 NVIDIA GTX1080Ti GPUs.…”

“…We also applied histopathological-oriented data augmentation. [ 5 ] The slide-level prediction was defined as the average of the top 100 probabilities of the pixel-level predictions.…”