ENAS-B: Combining ENAS with Bayesian Optimisation for Automatic Design of Optimal CNN Architectures for Breast Lesion Classification from Ultrasound Images

Ahmad, Mohammed and Du, Hongbo and AlZoubi, Alaa (2023) ENAS-B: Combining ENAS with Bayesian Optimisation for Automatic Design of Optimal CNN Architectures for Breast Lesion Classification from Ultrasound Images. Ultrasound Imaging. ISSN 0161-7346 (In Press)

Text UIX-23-0006 ENAS-B.pdf - Accepted Version Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (751kB)

Abstract

Efficient Neural Architecture Search (ENAS) is a recent development in searching for optimal cell structures for Convolutional Neural Network (CNN) design. It has been successfully used in various applications including ultrasound image classification for breast lesions. However, the existing ENAS approach only optimises cell structures rather than the whole CNN architecture nor its trainable hyperparameters. This paper presents a novel framework for automatic design of CNN architectures by combining strengths of ENAS and Bayesian Optimisation in two folds. Firstly, we use ENAS to search for optimal normal and reduction cells. Secondly, with the optimal cells and a suitable hyperparameter search space, we adopt Bayesian Optimisation to find the optimal depth of the network and optimal configuration of the trainable hyperparameters. To test the validity of the proposed framework, a dataset of 1,522 breast lesion ultrasound images is used for the searching and modelling. We then evaluate the robustness of the proposed approach by testing the optimized CNN model on three external datasets consisting of 727 benign and 506 malignant lesion images. We further compare the CNN model with the default ENAS-based CNN model, and then with CNN models based on the state-of-the-art architectures. The results (error rate of no more than 20.6% on internal tests and 17.3% on average of external tests) showed that the proposed framework generates robust and light CNN models.

Actions (login required)

View Item