Building Efficient CNN Architectures for Histopathology Images Analysis: A Case-Study in Tumor-Infiltrating Lymphocytes Classification

Abstract

BackgroundDeep learning methods have demonstrated remarkable performance in pathology image analysis, but they are computationally very demanding. The aim of our study is to reduce their computational cost to enable their use with large tissue image datasets.MethodsWe propose a method called Network Auto-Reduction (NAR) that simplifies a Convolutional Neural Network (CNN) by reducing the network to minimize the computational cost of doing a prediction. NAR performs a compound scaling in which the width, depth, and resolution dimensions of the network are reduced together to maintain a balance among them in the resulting simplified network. We compare our method with a state-of-the-art solution called ResRep. The evaluation is carried out with popular CNN architectures and a real-world application that identifies distributions of tumor-infiltrating lymphocytes in tissue images.ResultsThe experimental results show that both ResRep and NAR are able to generate simplified, more efficient versions of ResNet50 V2. The simplified versions by ResRep and NAR require 1.32× and 3.26× fewer floating-point operations (FLOPs), respectively, than the original network without a loss in classification power as measured by the Area under the Curve (AUC) metric. When applied to a deeper and more computationally expensive network, Inception V4, NAR is able to generate a version that requires 4× lower than the original version with the same AUC performance.ConclusionsNAR is able to achieve substantial reductions in the execution cost of two popular CNN architectures, while resulting in small or no loss in model accuracy. Such cost savings can significantly improve the use of deep learning methods in digital pathology. They can enable studies with larger tissue image datasets and facilitate the use of less expensive and more accessible graphics processing units (GPUs), thus reducing the computing costs of a study.

Effective and efficient active learning for deep learning-based tissue image analysis

“…We propose a strategy that replaces the target CNN with a less resource demanding or compact CNN to be used in the AL loop. Our approach automatically creates a simplified CNN via a process referred to as Network Auto-reduction ( NAR ) ( Meirelles et al 2022a ). NAR simplifies the target CNN in a systematic manner by reducing the number of layers, channels, and tensor resolutions.…”

“…In this work, to address this limitation, we incorporate a method, called network auto-reduction (NAR), in the AL process. NAR is an approach for simplification of CNNs we developed ( Meirelles et al 2022a ). The AL process with NAR reduces overall AL execution time by using simplified (hence, less computationally expensive) versions of the target networks in the sample selection step.…”

Spatial transcriptome reveals histology-correlated immune signature learnt by deep learning attention mechanism on H&E–stained images for ovarian cancer prognosis

“…Recent advancements in machine-learning algorithms for computer vision have created an interest in their applicability in digital pathology (13,14). Deep learning models trained on 5 data such as histological and computed tomography images have been used to predict signaling activity, mutation and prognosis (15)(16)(17)(18). Although ovarian tumor histological images have been employed to predict patient prognosis, the mechanism of the prediction is not fully understood.…”