Zihao Wang Research

Name in Chinese: 王 子豪 (Wang Zihao)

[Publications] [Contact] [Bio] [Google Scholar]


I am a Research Fellow in the Athinoula A. Martinos Center for Biomedical Imaging, and a Postdoctoral Fellow at Harvard University advised by Dr. Ona Wu. I received my Ph.D. in Computer Science at Inria, under the supervision of Dr. Hervé Delingette. Before that, I hold positions at Philips Healthcare and GE Healthcare.


Research Theme:

I focus on AI research for Engineering and Healthcare. My work involves developing principled machine learning algorithms for healthcare applications, with a particular emphasis on medical imaging. This includes working with generative learning models to reduce artifacts in CT imaging, applying Bayesian generative frameworks for organ shape analysis, developing discriminative learning methods for anatomical landmarks detection, and using attention mechanisms in image registration. Beyond healthcare, my research extends to the theoretical development of novel learning algorithms, such as harnessing statistical features for generation process in diffusion models and introducing backward stochastic differential equations for score matching in diffusion models. Additionally, I’ve utilized AI solutions to tackle practical problems in electronic engineering. My goal is to drive impactful and groundbreaking research in machine learning, focusing on identifying crucial challenges in engineering and clinical domains that can be effectively addressed using computational methods, while fostering valuable collaborations across diverse fields.




I gave a talk for the Statistical Atlases and Computational Modeling of the Heart (STACOM) workshop at MICCAI, 24 October 2022, Online.

Posted 24 Oct 2022

I obtain the Prix d’excellence d’Université Côte d’Azur 2021.

Posted 03 Jan 2022

I gave a talk for ‘DALI: The MICCAI Workshop on Data Augmentation, Labeling, and Imperfections’, 1 October 2021, Online. The presentations will introduce our recent work of using one-shot learning for cochlea landmarks. In this talk, he will show how to use only one annotated cochlea CT volume for detecting hundreds of cochlea volumes landmarks. This research is a part of CIMPLE research project.

Posted 24 Dec 2021

Dec. 21: We are organizing a special issue of Frontiers in computer science with the topic of a few shot learning in computer vision or medical imaging analysis. Any topics related to those domains are welcome. https://lnkd.in/eMY9AG4p

Posted 01 Dec 2021

Zihao Wang’s paper ‘One-shot Learning Landmarks Detection’ was selected for the People’s Choice Award at MICCAI DALI 2021.

Posted 01 Dec 2021