Automated Vessel Tortuosity Measurement from Retinal Imaging
June 1, 2025
·
1 min read
What problem are we solving?
Vessel tortuosity is a useful biomarker in multiple retinal and systemic conditions, but measurement is often subjective, inconsistent, or requires specialized tools that are not easily reproducible.
What we are building
An automated, open, and reproducible pipeline to:
- preprocess images
- segment/extract vessels (or accept vessel masks)
- compute tortuosity metrics (multiple definitions supported)
- export publication-ready summaries and QC overlays
My role
Primary developer and designer, algorithm selection, implementation, validation strategy, and documentation for open-source release.
Current status
Core pipeline stable; ongoing refinement, validation, and packaging for public release.
Outputs
- Open-source repository (planned)
- Methods write-up and example datasets (planned)
Computational Ophthalmology
Retinal Imaging
Python
Quantitative Biomarkers
Artificial Intelligence
AI
Authors
Ehsan Misaghi
(he/him)
Clinician-Scientist Trainee
Ehsan Misaghi is an MD/PhD Candidate at the University of Alberta working at the intersection of ophthalmology, genetics, and artificial intelligence.
His research focuses on inherited retinal disease and genotype–phenotype correlations in ocular disease, with an emphasis on mechanistic insight and translational relevance.
Alongside research, he builds and evaluates practical AI tools for clinical and educational settings, and he leads medical AI education, research, and community-building through the AI in Medical Systems Society (AIMSS) and related initiatives.
His goal is to advance rigorous, clinically useful research and translate it into improved diagnostics, care pathways, and responsible innovation.