Mina Mahdian
Stony Brook University, NY·Medicine & Clinical Research
Dentists increasingly use cone beam computed tomography (CBCT), a 3D X-ray scan of the teeth, jaws and nearby structures. These scans often capture more than teeth, including hardened, calcium-filled spots in arteries of the head and neck.
- Fields
Medicine & Clinical Research
Biomedical Engineering
Computer Science
- Topics
cone beam computed tomography
dental imaging
arterial calcification
- Grants
- 1 known grant, through May 2027
What they work on
- Develop an automated method to segment arterial calcifications in dental CBCT scans using anatomy-driven self-pretraining, so detection stays reliable even though targets fill under 1% of the image and labeled data are limited.
- Test the hypothesis that the shape and texture of calcified regions predict cardiovascular disease, using radiomic analysis of 1,000 CBCT scans from Stony Brook School of Dental Medicine.
- Build machine learning risk models that combine subtle CBCT imaging features with traditional clinical risk factors, and test whether they stratify cardiovascular risk better than conventional clinical metrics.
More about the research
Such arterial calcifications are linked to a higher risk of heart attack and stroke. They are easy to miss for dentists unfamiliar with the complex anatomy, and no measurable standard exists for judging cardiovascular disease (CVD) risk from them on CBCT.
Why it matters
Many patients get CBCT scans for implants, orthodontics or root canals. An automated tool could let dental professionals flag hidden arterial calcifications and refer patients for medical care earlier, which may lower the risk of heart attacks and strokes. Measurable imaging features would also give clinicians an objective way to judge cardiovascular risk from scans already taken.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Closest labs
Xiaofeng Yang
Emory University, GA·Biomedical Engineering
Develop and refine a deep learning method that reconstructs real-time on-board volumetric images from very few two-dimensional projections for direct, markerless three-dimensional lung tumor tracking.
Kazlin Mason
University of Virginia, VA·Medicine & Clinical Research
Develop a deep learning framework that automatically segments velopharyngeal structures in 3D and measures them in 2D from MRI scans of children with velopharyngeal anomalies.
Hua Li
Washington University in St Louis, MO·Biomedical Engineering
Develop a learning-based prognostic model that predicts cervical cancer treatment response and outcome by integrating RNA biomarkers, radiologic and histologic images, and clinicopathologic data from about 600 patients at two institutions.
The school
Stony Brook University: 32 labs on PI Pages, $319M a year on research, 289 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.