MuTGAN: Simultaneous Segmentation and Quantification of Myocardial Infarction without Contrast Agents via Joint Adversarial Learning Chenchu Xu 1 , Lei Xu 2 , Gary Brahm 1 , Heye Zhang 3⋆ , Shuo Li 1⋆ 1 Western university, London ON, Canada 2 Beijing AnZhen Hospital, Beijing, China 3 Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China Abstract. Simultaneous segmentation and full quantification (estima- tion of all diagnostic indices) of the myocardial infarction (MI) area are crucial for early diagnosis and surgical planning. Current clinical meth- ods still suffer from high-risk, non-reproducibility and time-consumption issues. In this study, the multitask generative adversarial networks (MuT- GAN) is proposed as a contrast-free, stable and automatic clinical tool to segment and quantify MIs simultaneously. MuTGAN consists of gen- erator and discriminator modules and is implemented by three seamless connected networks: spatio-temporal feature extraction network compre- hensively learns the morphology and kinematic abnormalities of the left ventricle through a novel three-dimensional successive convolution; joint feature learning network learns the complementarity between segmenta- tion and quantification through innovative inter- and intra-skip connec- tion; task relatedness network learns the intrinsic pattern between tasks to increase the accuracy of estimations through creatively utilized adver- sarial learning. MuTGAN minimizes a generalized divergence to directly optimize the distribution of estimations by using the competition pro- cess, which achieves pixel segmentation and full quantification of MIs. Our proposed method yielded a pixel classification accuracy of 96.46%, and the mean absolute error of the MI centroid was 0.977 mm, from 140 clinical subjects. These results indicate the potential of our proposed method in aiding standardized MI assessments. 1 Introduction Simultaneous segmentation and quantification (including pixel segmentation and full quantification of all indices such as the infarct size, segment percentage, perimeter, centroid, major axis length, minor axis length and orientation ) of a myocardial infarction (MI) are crucial to clinical treatment of the MI [1]. It seg- ments MI to predict the recovery of dysfunctional segments in chronic ischemic heart diseases or to select therapeutic options; it estimates all indices that in- dicate the presence, location, and transmurality of acute and chronic MI. The combination can obtain all information required for a thorough understanding ⋆ Corresponding Author: Dr. Shuo Li (slishuo@gmail.com) and Dr. Heye Zhang (hy.zhang@siat.ac.cn)