Riaan Zoetmulder1, Efstratios Gavves2, Matthan Caan3, Henk Marquering4. 1. Biomedical Engineering and Physics, Amsterdam UMC, Location AMC, Meibergdreef 15, 1105 AZ Amsterdam, the Netherlands; University of Amsterdam, Science Park 904, 1098 XH Amsterdam, the Netherlands. Electronic address: r.zoetmulder@amsterdamumc.nl. 2. University of Amsterdam, Science Park 904, 1098 XH Amsterdam, the Netherlands. 3. Biomedical Engineering and Physics, Amsterdam UMC, Location AMC, Meibergdreef 15, 1105 AZ Amsterdam, the Netherlands. 4. Biomedical Engineering and Physics, Amsterdam UMC, Location AMC, Meibergdreef 15, 1105 AZ Amsterdam, the Netherlands; Radiology & Nuclear Medicine, Amsterdam UMC, Location AMC, Meibergdreef 15, 1105 AZ Amsterdam, the Netherlands.
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.
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.
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
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