At Mohamed bin Zayed University of Artificial Intelligence in Abu Dhabi, the first PhD graduate has taken the university’s mission of practical AI innovation into the heart of the UAE’s healthcare system. Numan Saeed completed his doctorate in 2024 and now works as a research scientist at MBZUAI, where he leads efforts to develop AI tools that directly address clinical challenges faced by hospitals across the country.
From doctoral research to hospital partnerships
Saeed’s doctoral work centered on deep learning models for medical imaging, with a particular emphasis on head and neck cancers. He conducted this research within MBZUAI’s BioMedIA group under Associate Professor Mohammad Yaqub. The focus was on building models capable of analyzing multiple imaging modalities simultaneously to improve tumor detection and disease progression insights.
Since graduating, Saeed has expanded these efforts through direct collaboration with UAE hospitals. Key partners include Sheikh Shakhbout Medical City, Cleveland Clinic Abu Dhabi, Burjeel, and Corniche Hospital. These partnerships go beyond data sharing; they involve ongoing clinical feedback and expert annotations that ensure models perform reliably in real-world settings.
Building large-scale medical datasets
One of the most significant outcomes of these collaborations is the creation of extensive, locally relevant datasets. A standout example is a collection of more than 200,000 fetal ultrasound scans. Such scale is critical for training robust AI systems that can identify subtle patterns invisible to the human eye in many cases.
These datasets support applications ranging from fetal health monitoring to echocardiography. By training on UAE-specific data, the models better reflect local patient demographics and clinical practices, reducing the risk of performance gaps that often arise when using international datasets alone.
Cancer-focused foundation models and the HECKTOR project
Saeed’s team is actively developing a Cancer Foundation Model supported by a Department of Health grant. The project targets breast, colorectal, and head and neck cancers using multimodal data that includes electronic health records, imaging, and genomics.
A related initiative is the HECKTOR project, which leverages a large-scale, multi-modal dataset from more than 10 centers worldwide, including UAE sites. The work focuses on tumor segmentation from CT and PET scans, as well as staging and prognosis. Multiple radiologists across the UAE contribute to annotation, testing, and iterative feedback, ensuring the resulting lightweight, deployable solutions meet clinical standards.
Photo by Charles DeLoye on Unsplash
Developing medical world models for proactive care
Beyond pattern recognition, Saeed and colleagues are exploring medical world models. These systems simulate the dynamics of clinical environments, allowing clinicians to visualize plausible future trajectories for patients under different treatment scenarios.
The goal is to shift from passive prediction to proactive simulation. Doctors can compare potential outcomes—such as changes in disease progression or key health metrics—before selecting an intervention. This approach keeps human expertise central while augmenting decision-making with computational foresight.
Aligning AI research with real clinical needs
A recurring theme in Saeed’s work is the importance of closing the gap between technological development and day-to-day clinical priorities. He emphasizes that researchers should engage doctors early and often, rather than assuming which problems need solving.
This collaborative philosophy has shaped every stage of his post-PhD projects. Hospitals provide not only data but also the contextual understanding required to translate algorithms into tools that genuinely support workflows and improve patient outcomes.
MBZUAI’s role in UAE healthcare innovation
MBZUAI’s graduate programs in computer vision, machine learning, and computational biology provide the foundation for this type of translational research. The university’s emphasis on real-world impact aligns with national priorities outlined in UAE healthcare strategies, which increasingly integrate AI for earlier diagnosis and personalized medicine.
Partnerships with regulatory bodies such as the Department of Health – Abu Dhabi further strengthen these efforts, creating pathways for approved projects that can move from research prototypes to clinical deployment.
Implications for PhD training and future researchers
Saeed’s trajectory offers a model for current and prospective MBZUAI students. He advises choosing a research problem of genuine personal interest, maintaining persistence through challenges, and selecting a supervisor with strong alignment on goals.
His experience highlights the value of interdisciplinary collaboration—spanning computer science, clinical medicine, and data governance—as essential for success in medical AI. Students interested in similar paths can explore MBZUAI’s PhD programs in computer vision, machine learning, and the newly launched computational biology track.
Looking ahead: expanding precision medicine in the UAE
The work led by Saeed and his colleagues contributes to broader UAE ambitions in precision medicine. By developing models trained on local data and validated through hospital partnerships, these initiatives support earlier detection of common cancers and more tailored treatment strategies.
Future directions include further integration of multimodal data sources and continued refinement of world models that simulate clinical trajectories. These advancements position MBZUAI graduates as key contributors to the UAE’s evolving healthcare landscape.
Opportunities for academics and researchers
MBZUAI continues to recruit faculty and researchers focused on healthcare applications of AI. Those with expertise in medical imaging, trustworthy AI, or clinical data science will find a supportive environment for impactful work that directly benefits UAE hospitals and patients.
Prospective students and early-career researchers can review current openings and program details on the university’s site to explore pathways into this growing field.