A tool for automated retinal vascular morphology quantification and its applications #

Yukun Zhou, Daniel Alexander, Pearse Keane

14:30 Monday in 4Q56.

Part of the Mathematical and computational ophthalmology session.

Abstract #

We present AutoMorph, a tool for automated analysis of retinal vascular morphology on fundus photographs, to facilitate widespread research in ophthalmic and systemic diseases. AutoMorph consists of four functional modules: image preprocessing, image quality grading, anatomical segmentation (including binary vessel, artery/vein, and optic disc/cup segmentation), and vascular morphology feature measurement. Image quality grading and anatomical segmentation use the most recent deep learning techniques and ensemble strategies to achieve robust results. We externally validate the performance of each module on several independent publicly available datasets. The vascular morphology features measured from the AutoMorph segmentation map and expert annotation show good to excellent agreement even when external validation data show domain differences from training data (e.g., with different imaging devices). We also show recent application cases of AutoMorph in exploring the association between eye conditions and systemic diseases.