Mohammad Taha Parsayan

What is the title of your thesis?

PET/MRI Data Analysis and AI-Driven Classification of Alzheimer’s Disease Using Multimodal Data.

At which department and/or research unit did you complete your PhD?

Research Unit of Neurology, Department of Clinical Research, Faculty of Health Sciences, University of Southern Denmark, and Department of Neurology, Odense University Hospital.

Who was your principal supervisor?

Associate Professor Sasan Andalib.

What question did you aim to answer with your thesis?

Can AI using human brain imaging and clinical data help with the diagnosis of Alzheimer’s disease and its early stage?

What did you find?

AI models (deep learning) can accurately distinguish Alzheimer’s disease, mild cognitive impairment, and healthy individuals with over 80% accuracy and sensitivity.

In addition, the way brain structural and functional images are processed matters. Although many researchers use different methods to calculate brain activity (SUVR) from functional imaging, using cerebellum gray matter as a reference region gives more reliable and reproducible results.

How did you do it?

I analyzed human brain scans (PET and MRI) together with clinical data from patients. Then I designed and trained AI models (deep learning) to learn patterns in the data and classify patients into different groups.

I also compared different ways of measuring brain activity (SUVR) to find the most reliable method.

How can your research be applied (in the clinic, society, etc.)?

It can work as a support tool to help doctors diagnose Alzheimer’s earlier and more accurately. This supports better patient care, earlier treatment planning, and improved monitoring of disease progression.