Cheng Ouyang
Cheng Ouyang is a Departmental Lecturer on Oxford's Turing AI Fellowship team and a Visiting Researcher with Imperial's BioMedIA group. At Oxford, he develops data-efficient, robust machine learning for medical image and signal computing. He completed a PhD in computing at Imperial in 2023, worked on cardiac imaging at Imperial's Institute of Clinical Sciences, and joined Oxford's Biomedical Image Analysis Lab in April 2024. His papers include PnP-AdaNet for cross-modality cardiac segmentation, a superpixel method for few-shot medical image segmentation, a topology-aware MICCAI paper, and MedVLM-R1, a medical vision-language model that uses reinforcement learning to generate human-interpretable reasoning. He won MICCAI challenges for cardiac MR segmentation in 2019 and fetal brain annotation in 2022.
- Now at
- University of Oxford and Imperial College London
Timeline
-
2018
PnP-AdaNet for cardiac segmentation
In 2018, he coauthors PnP-AdaNet, a plug-and-play adversarial domain adaptation network with a benchmark for cross-modality cardiac segmentation.
-
2019
A MICCAI cardiac MR win
At MICCAI 2019, he wins the Multi-sequence Cardiac MR Segmentation Challenge.
-
2020
Few-shot segmentation with superpixels
In 2020, he coauthors Self-Supervision with Superpixels, a method for training few-shot medical image segmentation without annotation.
-
2022
A fetal brain challenge win
In 2022, he wins the Multi-Center Fetal Brain Tissue Annotation Challenge (FeTA) at MICCAI.
-
2023
PhD and topology-aware segmentation
In 2023, he earns a PhD in Computing at Imperial College London and coauthors Robust Segmentation via Topology Violation Detection and Feature Synthesis for MICCAI 2023, a paper on topology-aware medical image segmentation.
-
April 2024
A lectureship at Oxford
In April 2024, he joins Oxford as a Departmental Lecturer in AI for Healthcare Imaging, working in the Biomedical Image Analysis Lab. Before joining Oxford, he works as a cardiac imaging postdoctoral researcher at Imperial's Institute of Clinical Sciences.
-
2025
MedVLM-R1 and reinforcement learning
In 2025, he coauthors MedVLM-R1, a medical vision-language model that uses reinforcement learning to discover human-interpretable reasoning paths without reasoning references.
-
As of October 2026
Departmental Lecturer at Oxford
As of October 2026, he is a Departmental Lecturer on Oxford's Turing AI Fellowship team and a Visiting Researcher with Imperial's BioMedIA group. At Oxford, his research focuses on data-efficient, robust machine learning for medical image and signal computing.
Career
- Since 2024University of OxfordDepartmental Lecturer in AI for Healthcare Imaging
- NowImperial College LondonVisiting Researcher
Education
- 2023Imperial College LondonPhD in Computing
Colleagues
Sources and credits
Photo: We have no published photo of Cheng Ouyang yet. Send us a free-licensed photo.
- PnP-AdaNet: Plug-and-Play Adversarial Domain Adaptation Network with a Benchmark at Cross-modality Cardiac Segmentation
- Profile for Dr Cheng Ouyang, Departmental Lecturer in the Department of Engineering Science, University of Oxford
- Self-Supervision with Superpixels: Training Few-shot Medical Image Segmentation without Annotation
- Welcome to our new Departmental Lecturers and Associate Professors - news from the Department of Engineering Science, University of Oxford
- Robust Segmentation via Topology Violation Detection and Feature Synthesis - Technical University of Munich
- MedVLM-R1: Incentivizing Medical Reasoning Capability of Vision-Language Models (VLMs) via Reinforcement Learning
- The Turing AI Fellowship Team at the University of Oxford
- Cheng Ouyang – BioMedIA