Literature DB >> 34875512

Domain- and task-specific transfer learning for medical segmentation tasks.

Riaan Zoetmulder1, Efstratios Gavves2, Matthan Caan3, Henk Marquering4.   

Abstract

BACKGROUND AND OBJECTIVES: Transfer learning is a valuable approach to perform medical image segmentation in settings with limited cases available for training convolutional neural networks (CNN). Both the source task and the source domain influence transfer learning performance on a given target medical image segmentation task. This study aims to assess transfer learning-based medical segmentation task performance for various source task and domain combinations.
METHODS: CNNs were pre-trained on classification, segmentation, and self-supervised tasks on two domains: natural images and T1 brain MRI. Next, these CNNs were fine-tuned on three target T1 brain MRI segmentation tasks: stroke lesion, MS lesions, and brain anatomy segmentation. In all experiments, the CNN architecture and transfer learning strategy were the same. The segmentation accuracy on all target tasks was evaluated using the mIOU or Dice coefficients. The detection accuracy was evaluated for the stroke and MS lesion target tasks only.
RESULTS: CNNs pre-trained on a segmentation task on the same domain as the target tasks resulted in higher or similar segmentation accuracy compared to other source task and domain combinations. Pre-training a CNN on ImageNet resulted in a comparable, but not consistently higher lesion detection rate, despite the amount of training data used being 10 times larger.
CONCLUSIONS: This study suggests that optimal transfer learning for medical segmentation is achieved with a similar task and domain for pre-training. As a result, CNNs can be effectively pre-trained on smaller datasets by selecting a source domain and task similar to the target domain and task.
Copyright © 2021 Elsevier B.V. All rights reserved.

Entities:  

Keywords:  Deep learning; Domain adaptation; MRI; Task transfer learning

Mesh:

Year:  2021        PMID: 34875512     DOI: 10.1016/j.cmpb.2021.106539

Source DB:  PubMed          Journal:  Comput Methods Programs Biomed        ISSN: 0169-2607            Impact factor:   5.428


  3 in total

1.  Medical Image Classification Using Transfer Learning and Chaos Game Optimization on the Internet of Medical Things.

Authors:  Alhassan Mabrouk; Abdelghani Dahou; Mohamed Abd Elaziz; Rebeca P Díaz Redondo; Mohammed Kayed
Journal:  Comput Intell Neurosci       Date:  2022-07-13

2.  Artificial Intelligence for Colonoscopy: Past, Present, and Future.

Authors:  Wallapak Tavanapong; JungHwan Oh; Michael A Riegler; Mohammed Khaleel; Bhuvan Mittal; Piet C de Groen
Journal:  IEEE J Biomed Health Inform       Date:  2022-08-11       Impact factor: 7.021

3.  Global Research Trends of Artificial Intelligence on Histopathological Images: A 20-Year Bibliometric Analysis.

Authors:  Wentong Zhou; Ziheng Deng; Yong Liu; Hui Shen; Hongwen Deng; Hongmei Xiao
Journal:  Int J Environ Res Public Health       Date:  2022-09-15       Impact factor: 4.614

  3 in total

北京卡尤迪生物科技股份有限公司 © 2022-2023.