Subtype-aware Unsupervised Domain Adaptation for Medical Diagnosis

Authors

  • Xiaofeng Liu Dept. of Radiology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA Dept. of Neurology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA
  • Xiongchang Liu Dept. of Neurology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA China University of Mining and Technology, China
  • Bo Hu Dept. of Neurology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA Beijing University of Posts and Telecommunications, China
  • Wenxuan Ji School of Artificial Intelligence, Nankai University, China
  • Fangxu Xing Dept. of Radiology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA
  • Jun Lu Dept. of Neurology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA
  • Jane You Dept. of Computing, The Hong Kong Polytechnic University, Hong Kong
  • C.-C. Jay Kuo Dept. of Electrical and Computer Engineering, University of Southern California, Los Angeles, CA, USA
  • Georges El Fakhri Dept. of Radiology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA
  • Jonghye Woo Dept. of Radiology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA

Keywords:

Applications, Healthcare, Medicine & Wellness, Classification and Regression, Clustering

Abstract

Recent advances in unsupervised domain adaptation (UDA) show that transferable prototypical learning presents a powerful means for class conditional alignment, which encourages the closeness of cross-domain class centroids. However, the cross-domain inner-class compactness and the underlying fine-grained subtype structure remained largely underexplored. In this work, we propose to adaptively carry out the fine-grained subtype-aware alignment by explicitly enforcing the class-wise separation and subtype-wise compactness with intermediate pseudo labels. Our key insight is that the unlabeled subtypes of a class can be divergent to one another with different conditional and label shifts, while inheriting the local proximity within a subtype. The cases with or without the prior information on subtype numbers are investigated to discover the underlying subtype structure in an online fashion. The proposed subtype-aware dynamic UDA achieves promising results on a medical diagnosis task.

Downloads

Published

2021-05-18

How to Cite

Liu, X., Liu, X., Hu, B., Ji, W., Xing, F., Lu, J., You, J., Kuo, C.-C. J., El Fakhri, G., & Woo, J. (2021). Subtype-aware Unsupervised Domain Adaptation for Medical Diagnosis. Proceedings of the AAAI Conference on Artificial Intelligence, 35(3), 2189-2197. Retrieved from https://ojs.aaai.org/index.php/AAAI/article/view/16317

Issue

Section

AAAI Technical Track on Computer Vision II