Transfer Learning for Domain Adaptation in MRI: Application in Brain Lesion Segmentation

By Mohsen Ghafoorian, Alireza Mehrtash, Tina Kapur, Nico Karssemeijer, Elena Marchiori, Mehran Pesteie, C. Guttmann, F.‐E. Leeuw, Clare M. Tempany, B. van Ginneken and 4 more. International Conference on Medical Image Computing and Computer-Assisted Intervention, 2017

391 citations recorded by the source corpus.

Magnetic Resonance Imaging (MRI) is widely used in routine clinical diagnosis and treatment. However, variations in MRI acquisition protocols result in different appearances of normal and diseased tissue in the images. Convolutional neural networks (CNNs), which have shown to be successful in many medical image analysis tasks, are typically sensitive to the variations in imaging protocols. Therefore, in many cases, networks trained on data acquired with one MRI protocol, do not perform satisfactorily on data acquired with different protocols. This limits the use of models trained with large annotated legacy datasets on a new dataset with a different domain which is often a recurring situation in clinical settings. In this study, we aim to answer the following central questions regarding domain adaptation in medical image analysis: Given a fitted legacy model, (1) How much data from the new domain is required for a decent adaptation of the original network?; and, (2) What portion of the pre-trained model parameters should be retrained given a certain number of the new domain training samples? To address these questions, we conducted extensive experiments in white matter hyperintensity segmentation task. We trained a CNN on legacy MR images of brain and evaluated the performance of the domain-adapted network on the same task with images from a different domain. We then compared the performance of the model to the surrogate scenarios where either the same trained network is used or a new network is trained from scratch on the new dataset. The domain-adapted network tuned only by two training examples achieved a Dice score of 0.63 substantially outperforming a similar network trained on the same set of examples from scratch.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/102a580a89b7e9d5a6d8e79864db183d6b5fb6ad

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. P. Heng
2. Qi-Yan Dou
3. Jing Qin
4. Hao Chen
5. Bo Du
6. Cheng Chen
7. Pingkun Yan
8. Qikui Zhu

In Computer Vision and Pattern Recognition

1. Holger R. Roth
2. P. Heng
3. Daguang Xu
4. Qi-Yan Dou
5. Dong Yang
6. Hao Chen
7. Jing Qin
8. Bo Du

Citations in the corpus

• An introduction to domain adaptation and transfer learning (2018, background): “Unfortunately, due to the mechanical configuration, calibration, vendor and acquisition protocol of the scanner, the images it produces will differ from images produced by other scanners [196,79,123].”
• Synergistic Image and Feature Adaptation: Towards Cross-Modality Domain Adaptation for Medical Image Segmentation (2019, methodology): “To recover model performance, an easy way is to re-train or fine-tune models with additional labeled data from the target domain (Van Opbroek et al. 2015; Ghafoorian et al. 2017).”
• A Review of Deep Learning on Medical Image Analysis (2020, methodology): “[32] 2017 Brain lesion Feature extractor”
• Unsupervised Bidirectional Cross-Modality Adaptation via Deeply Synergistic Image and Feature Alignment for Medical Image Segmentation (2020, background): “source data with additional labeled target data [4].”
• Generalizing Deep Learning for Medical Image Segmentation to Unseen Domains via Deep Stacked Transformation (2020, background): “The first, transfer learning, is the process of fine-tuning a portion of a pre-trained network (usually the last few layers [5] or shared convolutional filters [6]).”
• Domain Adaptation for Medical Image Analysis: A Survey (2021, neutral): “Each group is further divided into three categories, i.e. , supervised, semi-supervised and unsupervised methods.”
• Confidence Calibration and Predictive Uncertainty Estimation for Deep Medical Image Segmentation (2019, background): “to detect out-of-distribution inputs is specifically important for medical imaging applications as deep networks are sensitive to domain shift, which is a recurring situation in medical imaging [15].”
• SynthSeg: Segmentation of brain MRI scans of any contrast and resolution without retraining (2021, neutral): “While these networks obtain fast and accurate re-sults on their training domain, they do not generalise well to unseen contrasts [45] and resolutions [28], an issue known as the “domain-gap” problem [54].”
• Boundary-weighted Domain Adaptive Neural Network for Prostate MR Image Segmentation (2019, methodology): “[25] conducted extensive experiments in white matter hyperintensity segmentation and evaluated the”
• Transfer learning: a friendly introduction (2022, background): “adaptation process tries to alter the source domain to bring closer the distribution of the source domain to the target domain [23].”
• Secure and Robust Machine Learning for Healthcare: A Survey (2020, methodology): “image analysis tasks such as MRI segmentation [136], [193], chest X-ray classification [194], and multi-class Alzheimer disease classification [195], etc.”