Transfer Learning for Image Segmentation by Combining Image Weighting and Kernel Learning

TY - JOUR

T1 - Transfer Learning for Image Segmentation by Combining Image Weighting and Kernel Learning

AU - van Opbroek, Annegreet

AU - Achterberg, Hakim C.

AU - Vernooij, Meike W.

AU - de Bruijne, Marleen

PY - 2019/1

Y1 - 2019/1

N2 - Many medical image segmentation methods are based on the supervised classification of voxels. Such methods generally perform well when provided with a training set that is representative of the test images to the segment. However, problems may arise when training and test data follow different distributions, for example, due to differences in scanners, scanning protocols, or patient groups. Under such conditions, weighting training images according to distribution similarity have been shown to greatly improve performance. However, this assumes that a part of the training data is representative of the test data; it does not make unrepresentative data more similar. We, therefore, investigate kernel learning as a way to reduce differences between training and test data and explore the added value of kernel learning for image weighting. We also propose a new image weighting method that minimizes maximum mean discrepancy (MMD) between training and test data, which enables the joint optimization of image weights and kernel. Experiments on brain tissue, white matter lesion, and hippocampus segmentation show that both kernel learning and image weighting, when used separately, greatly improve performance on heterogeneous data. Here, MMD weighting obtains similar performance to previously proposed image weighting methods. Combining image weighting and kernel learning, optimized either individually or jointly, can give a small additional improvement in performance.

AB - Many medical image segmentation methods are based on the supervised classification of voxels. Such methods generally perform well when provided with a training set that is representative of the test images to the segment. However, problems may arise when training and test data follow different distributions, for example, due to differences in scanners, scanning protocols, or patient groups. Under such conditions, weighting training images according to distribution similarity have been shown to greatly improve performance. However, this assumes that a part of the training data is representative of the test data; it does not make unrepresentative data more similar. We, therefore, investigate kernel learning as a way to reduce differences between training and test data and explore the added value of kernel learning for image weighting. We also propose a new image weighting method that minimizes maximum mean discrepancy (MMD) between training and test data, which enables the joint optimization of image weights and kernel. Experiments on brain tissue, white matter lesion, and hippocampus segmentation show that both kernel learning and image weighting, when used separately, greatly improve performance on heterogeneous data. Here, MMD weighting obtains similar performance to previously proposed image weighting methods. Combining image weighting and kernel learning, optimized either individually or jointly, can give a small additional improvement in performance.

UR - http://www.scopus.com/inward/record.url?scp=85050597954&partnerID=8YFLogxK

U2 - 10.1109/TMI.2018.2859478

DO - 10.1109/TMI.2018.2859478

M3 - Journal article

C2 - 30047874

AN - SCOPUS:85050597954

SN - 0278-0062

VL - 38

SP - 213

EP - 224

JO - IEEE Transactions on Medical Imaging

JF - IEEE Transactions on Medical Imaging

IS - 1

M1 - 8419778

ER -