MedNAS: Multiscale Training-Free Neural Architecture Search for Medical Image Analysis
IEEE Transactions on Evolutionary Computation · 2024-06 · 28 (3) · 668-681
DOI: 10.1109/tevc.2024.3352641 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Evolutionary Computation, 2024-06. Verfasst von Yan Wang und 12 weiteren Autor:innen. Die Arbeit wurde bisher 44-mal zitiert (Crossref) und nennt 61 Literaturangaben. Gefördert von: National Research Foundation of Singapore under its AI Singapore Programme, Agency for Science, Technology and Research (A*STAR) through its AME Programmatic Funding Scheme.
Abstract (Originalsprache)
Deep neural networks have demonstrated impressive results in medical image analysis, but designing suitable architectures for each specific task is expertise dependent and time consuming. Neural architecture search (NAS) offers an effective means of discovering architectures. It has been highly successful in numerous applications, particularly in natural image classification. Yet, medical images possess unique characteristics, such as small regions and a wide variety of lesion sizes, that differentiate them from natural images. Furthermore, most current NAS methods struggle with high computational costs, especially when dealing with high-resolution image datasets. In this article, we present a novel evolutionary NAS method called multiscale training-free neural architecture search (MSTF-NAS) to address these challenges. Specifically, to accommodate the broad range of lesion region sizes in disease diagnosis, we develop a new reduction cell search space that enables the search algorithm to explicitly identify the optimal scale combination for multiscale feature extraction. To overcome the issue of high computational costs, we utilize training-free indicators as performance measures for candidate architectures, which allows us to search for the optimal architecture more efficiently. More specifically, by considering the capability and simplicity of various networks, we formulate a multiobjective optimization problem that involves two training-free indicators and model complexity for candidate architectures. Extensive experiments on a large medical image benchmark and a publicly available breast cancer detection dataset are conducted. The empirical results demonstrate that our MSTF-NAS outperforms both human-designed architectures and current state-of-the-art NAS algorithms on both datasets, indicating the effectiveness of our proposed method.
Förderung
- National Research Foundation of Singapore under its AI Singapore Programme
- Agency for Science, Technology and Research (A*STAR) through its AME Programmatic Funding Scheme
Zitieren
APA
Wang, Y., Zhen, L., Zhang, J., Li, M., Zhang, L., Wang, Z., Feng, Y., Xue, Y., Wang, X., Chen, Z., Luo, T., Goh, R. S. M., & Liu, Y. (2024). MedNAS: Multiscale Training-Free Neural Architecture Search for Medical Image Analysis. IEEE Transactions on Evolutionary Computation, 28(3), 668-681. https://doi.org/10.1109/tevc.2024.3352641
Vancouver
Wang Y, Zhen L, Zhang J, Li M, Zhang L, Wang Z, et al. MedNAS: Multiscale Training-Free Neural Architecture Search for Medical Image Analysis. IEEE Transactions on Evolutionary Computation. 2024;28(3):668-681. doi:10.1109/tevc.2024.3352641
BibTeX
@article{Wang2024,
author = {Wang, Yan and Zhen, Liangli and Zhang, Jianwei and Li, Miqing and Zhang, Lei and Wang, Zizhou and Feng, Yangqin and Xue, Yu and Wang, Xiao and Chen, Zheng and Luo, Tao and Goh, Rick Siow Mong and Liu, Yong},
title = {MedNAS: Multiscale Training-Free Neural Architecture Search for Medical Image Analysis},
journal = {IEEE Transactions on Evolutionary Computation},
year = {2024},
volume = {28},
number = {3},
pages = {668--681},
doi = {10.1109/tevc.2024.3352641}
}
EUVIMED