Dynamic Learning Rate in Deep CNN Model for Metastasis Detection and Classification of Histopathology Images.

Anil Johny,K. N. Madhusoodanan,Markos G. Tsipouras

doi:10.1155/2021/5557168

Anil Johny, K. N. Madhusoodanan + Show 1 more

Open Access

https://doi.org/10.1155/2021/5557168

Copy DOI

Abstract

Diagnosis of different breast cancer stages using histopathology whole slide images (WSI) is the gold standard in determining the grade of tissue metastasis. Computer-aided diagnosis (CAD) assists medical experts as a second opinion tool in early detection to prevent further proliferation. The field of pathology has advanced so rapidly that it is possible to obtain high-quality images from glass slides. Patches from the region of interest in histopathology images are extracted and trained using artificial neural network models. The trained model primarily analyzes and predicts the histology images for the benign or malignant class to which it belongs. Classification of medical images focuses on the training of models with layers of abstraction to distinguish between these two classes with less false-positive rates. The learning rate is the crucial hyperparameter used during the training of deep convolutional neural networks (DCNN) to improve model accuracy. This work emphasizes the relevance of the dynamic learning rate than the fixed learning rate during the training of networks. The dynamic learning rate varies with preset conditions between the lower and upper boundaries and repeats at different iterations. The performance of the model thus improves and attains comparatively high accuracy with fewer iterations.

Highlights

Deep learning has emerged as a state-of-the-art technology in computer vision and speech recognition in recent years
A custom convolutional neural network (CNN) model is designed and trained using a dynamic learning rate to improve the performance of the network for the classification of histology images
The learning rate is the crucial hyperparameter which decides the quality of CNN model training as it imparts fine-tuning in classification tasks

Summary

Introduction

Deep learning has emerged as a state-of-the-art technology in computer vision and speech recognition in recent years. CNN can learn spatial features in medical images adaptively using the backpropagation algorithm. Metastasis detection with the assistance of the algorithm requires training the model with adequate images so that the model learns characteristic features in the spatial domain. Deep learning methods are effective [3] when the number of available images is large during the training stage. The network learns from the given dataset by updating the weights after each training step for the given number of classes and classifies images by reducing training loss. Optimization of the deep neural network (DNN) model involves fine-tuning of hyperparameters like the learning rate, batch size (BS), and momentum to improve model performance in taskspecific applications. Model training with a fixed learning rate strategy is used than its counterparts.

Objectives

Methods

Results

Discussion

Conclusion