Homepage - Imran Razzak

About Me

I am an Associate Professor of Computational Biology at Mohamed bin Zayed University of Artificial Intelligence (MBZUAI), where I lead the GenMI research lab. I am also CEO of MedOS and MedOmni. MedOS builds AI for medicine, while MedOmni focuses on evidence-based medicine. Previously, I worked in human-centered machine learning at UNSW and computer science at Deakin University.

My research connects personalized medicine, medical imaging, and multimodal clinical AI. I develop models that bring together imaging, language, and other biomedical data for clinical reasoning and spatial grounding, with an emphasis on robust, trustworthy systems and longitudinal healthcare.

Open positions: I welcome prospective PhD students, postdoctoral researchers, visiting students, and remote interns interested in medical foundation models and clinical reasoning. Please read the application guidance and email your CV and a short research proposal.

Featured Research

MedMO

Medical image understanding, spatial grounding and report generation in one multimodal model.

Paper · Code · Models

CARL

Representation learning designed to preserve causal structure.

Paper

Oral microbiome

Population-scale investigation of links between the oral microbiome and metabolic health.

Paper

Recent Publications

2025–2026 highlights · IEEE TIP, IEEE JBHI, CVPR, ICLR and Nature Communications

Fig. 2 · source paper

SAM-aware Test-time Adaptation for Universal Medical Image Segmentation

Jianghao Wu, Yicheng Wu, Yutong Xie, Wenjia Bai, You Zhang, Feilong Tang, Yulong Li, Imran Razzak, Daniel F Schmidt, Yasmeen George

IEEE Transactions on Image Processing (TIP) — Accepted, September 2026

[Google Scholar] [Paper] [arXiv]

Fig. 4 · source paper

Attention-Driven Framework for Non-Rigid Medical Image Registration

Muhammad Zafar Iqbal, Ghazanfar Farooq Siddiqui, Anwar Ul Haq, Imran Razzak

IEEE Journal of Biomedical and Health Informatics (JBHI) — Accepted, September 2026

[Google Scholar] [Paper] [arXiv]

Image supplied by Imran Razzak

Population-scale characterization of the oral microbiome and associations with metabolic health

H Xue, A Godneva, F Tang, H Li, Y Li, M Hu, R Li, J Su, E Segal, I Razzak

Nature Communications · 2026

[Google Scholar]

Image supplied by Imran Razzak

CARL: Preserving Causal Structure in Representation Learning

Y Li, X Liu, F Tang, Z Lu, M Hu, Y Li, H Xue, P Guo, J Su, Y Xie, E Segal, ...

ICLR · 2026

[Google Scholar]

Figure 2 · source paper

MedMO: Grounding and Understanding Multimodal Large Language Model for Medical Images

Ankan Deria, Komal Kumar, Adinath Madhavrao Dukre, Eran Segal, Salman Khan, Imran Razzak

CVPR · 2026

[Google Scholar] [Paper] [Code] [Models] [arXiv]

Figure 1 · source paper

LATA: Laplacian-Assisted Transductive Adaptation for Conformal Uncertainty in Medical VLMs

Behzad Bozorgtabar, Dwarikanath Mahapatra, Sudipta Roy, Muzammal Naseer, Imran Razzak, Zongyuan Ge

CVPR · 2026

[Google Scholar] [Paper] [arXiv]

Figure 1 · source paper

CHIPS: Efficient CLIP Adaptation via Curvature-aware Hybrid Influence-based Data Selection

Xinlin Zhuang, Yichen Li, Xiwei Liu, Haolin Yang, Yifan Lu, Ziyun Zou, Yulong Li, Huifa Li, Dongliang Chen, Qinglei Wang, Weiyang Liu, Ying Qian, Jiangming Shi, Imran Razzak

CVPR · 2026

[Google Scholar] [Paper] [arXiv]

Figure 2 · source paper

Towards Efficient Medical Reasoning with Minimal Fine-Tuning Data

Xinlin Zhuang, Feilong Tang, Haolin Yang, Xiwei Liu, Ming Hu, Huifa Li, Haochen Xue, Junjun He, Zongyuan Ge, Yichen Li, Ying Qian, Imran Razzak

CVPR · 2026

[Google Scholar] [Paper] [arXiv]

Figure 4 · source paper

Seeing Far and Clearly: Mitigating Hallucinations in MLLMs with Attention Causal Decoding

Feilong Tang, Chengzhi Liu, Zhongxing Xu, Ming Hu, Zelin Peng, Zhiwei Yang, Jionglong Su, Minquan Lin, Yifan Peng, Xuelian Cheng, Imran Razzak, Zongyuan Ge

CVPR · 2025

[Google Scholar] [Paper] [arXiv]

All publications

Research Demos

12-second walkthroughs of paper figures and results.