MD and research associate in abdominal radiology at Shariati Hospital, Tehran University of Medical Sciences. My work sits at the intersection of quantitative MRI, IBD imaging, and clinical AI.
I build reproducible tools and models for problems that matter clinically: stratifying disease activity from MRE features, extracting structured metrics from radiology reports, and making complex imaging workflows usable in research pipelines.
- Bayesian multilevel model for segmental Mayo Endoscopic Score triage in ulcerative colitis, from MRE features and fecal calprotectin
- Quantitative MRE in ulcerative colitis: motility, chronicity, and activity
- Building toward an agentic LLM pipeline for Lémann Index scoring from free-text MRE reports in Crohn's disease, currently at the deterministic foundation stage
- Two first-author oral presentations at ESGAR 2026
- Ongoing collaboration with UCL / Motilent on quantitative bowel MRE
- uc-mre-activity-risk-stratifier — Bayesian ordinal model for UC disease activity from MRE
- abdominal-radiology-reports — reproducible, privacy-preserving pipeline for structuring free-text radiology reports
- radiology-clip-mini — compact CLIP baseline for chest radiograph and report retrieval
- lemann-li-app — Streamlit calculator for the updated Lémann Index
- ibd-mre-multimodal — early-stage multimodal pipeline for IBD activity