Numan Saeed
Assistant Teaching Professor
Research Interests
Professor Saeed's teaching and research interests span medical imaging and multimodal machine learning, with a particular focus on foundation models that learn from the imaging data clinicians already acquire every day, and on world models that go further by learning how anatomy and disease change over time. His work covers vision-language and self-supervised models for ultrasound, PET/CT, and MRI, spanning fetal ultrasound, echocardiography, and oncology, and extends to knowledge distillation methods that compress large models for deployment on point-of-care and mobile devices.
Professor Saeed also works on multimodal survival analysis, combining imaging with electronic health records to predict patient outcomes. In the classroom, he teaches medical imaging and multimodal computer vision, including CV8501: Medical Multimodal Vision, and has taught Deep Learning, Machine Learning, Life-long Learning, and Medical Imaging: Physics and Analysis. He is especially interested in moving students from benchmark performance toward clinically meaningful evaluation, and in mentoring junior researchers toward independent scientific careers.
Email
Professor Saeed serves as principal organizer of the HECKTOR challenge at MICCAI, which provides the largest publicly available multimodal PET/CT head and neck cancer dataset with clinical outcomes and has attracted over 100 participating teams; he is currently coordinating the 2026 edition. He collaborates with GE Healthcare on vision-language models for automated detection of fetal cardiac abnormalities in ultrasound video, and holds a U.S. patent on deep learning for head and neck tumor segmentation and survival prediction. He was previously a Visiting Researcher at the Technical University of Munich, where he developed self-supervised methods for kinetic modeling of dynamic PET data. He serves as an Area Chair and reviewer for MICCAI, reviews for other leading venues in medical imaging and machine learning, and is a recipient of several MICCAI awards.
- Ph.D. in Machine Learning, Mohamed bin Zayed University of Artificial Intelligence (MBZUAI), UAE
- M.Sc. in Microsystems, Masdar Institute of Science and Technology (in collaboration with MIT), UAE
- B.Sc. in Electrical Engineering, National University of Computer and Emerging Sciences, Pakistan
- U.S. Patent (US 12,629,123 B2), "Deep learning apparatus and method for segmentation and survival prediction for head and neck tumors" (with I. Sobirov, R. Majzoub, M. Yaqub)
- MICCAI Grant Awards, 2022, 2024
- First Ph.D. graduate of MBZUAI, 2023
- MICCAI Student Travel Award, 2022
- First Place, HECKTOR Challenge, MICCAI, 2021
- Graduate Student Fellowship for Ph.D., MBZUAI, 2021
- Graduate Fellowship for M.Sc., Masdar Institute of Science and Technology, 2015
- Gold, Silver, and Bronze Medals across semesters; five-time Dean's List and Rector's List of Honors, National University of Computer and Emerging Sciences, 2009–2013
- Maani F.*, Saeed N.*, et al. "FetalCLIP: A Visual-Language Foundation Model for Fetal Ultrasound Image Analysis", npj Digital Medicine, 2026. (*equal contribution) – the first vision-language foundation model for fetal ultrasound, pretrained on 210,035 image–text pairs.
- Saeed N., Ridzuan M., et al.: "SurvRNC: Learning Ordered Representations for Survival Prediction using Rank-N-Contrast", MICCAI, 2024.
- Saeed N., Ridzuan M., et al.: "MGMT Promoter Methylation Status Prediction Using MRI Scans? An Extensive Experimental Evaluation of Deep Learning Models", Medical Image Analysis, 2023.
- Saeed N., Sobirov I., et al.: "TMSS: An End-to-End Transformer-Based Multimodal Network for Segmentation and Survival Prediction", MICCAI, 2022.
- Saeed N. (Principal Organizer), et al.: "The HECKTOR 2025 Challenge: Head and Neck Tumor Segmentation and Outcome Prediction in PET/CT", under review, Scientific Data.
- CardioBench. IJCAI-ECAI 2026, AI and Health Special Track.