Developing novel machine learning frameworks for medical neuroimaging — from probabilistic Gaussian processes to deep graph neural networks for MRI and DTI analysis.
My research advances clinical artificial intelligence by developing novel machine learning frameworks to analyze complex scientific data, with a primary focus on medical neuroimaging.
My career represents a logical progression from my Ph.D., where I built foundational ML models for geophysical data, to my current focus on medicine. The core challenge remains the same: extracting meaningful information from noisy, high-dimensional, and often incomplete data.
This transition is exemplified by my work evolving Gaussian Processes — originally used for mineral estimation — into a novel framework (MGP) to solve the critical problem of missing data in clinical neuroimaging.
At Weill Cornell Medicine, I focus on enhancing the reliability and clinical application of ML in medical imaging and healthcare data analysis, including deep graph neural networks for MRI/DTI, harmonization, and modeling trajectories from longitudinal EHR data.
Postdoctoral Research Associate
Dept. of Radiology, Weill Cornell Medicine
New York, USA — Jul 2024–Present
Probabilistic ML · Gaussian Processes · Deep Graph Neural Networks · Neonatal Neuroimaging · Longitudinal EHR
Ph.D. Mining Engineering — Isfahan University of Technology, 2018
M.Sc. — Isfahan University of Technology, 2012
B.Sc. — University of Tehran, 2009
ANECA accredited (Spain) · Peer reviewer for Neural Networks, MICCAI, Knowledge-Based Systems, and more
Full list available on Google Scholar.
Registered software for medical image analysis — neonatal MRI/US segmentation and biomarker extraction. Reg. No. 2211222681375, Spain, 2022.
Author and maintainer of 25+ public repositories on GitHub covering deep learning, Gaussian processes, and scientific computing.
