arXiv cs.AI· Sharan Sharmeela Karuppannan, Yoga Sri Varshan V, Subin Sahayam, Umarani Jayaraman·· 10 小时前AI 评分22
五阶段可解释流水线：从眼底图像估计 AVR 并筛查视网膜病变
A Stage-Wise Explainable Pipeline for Arteriolar-to-Venular Ratio Estimation and Retinopathy Screening
AI 导读
研究者提出一套五阶段可解释流水线，从眼底图像估计小动脉-小静脉直径比（AVR）并进行探索性视网膜病变筛查，每个阶段都输出可检查的中间结果。在 30 张 HRF 图像的探索性队列上，该流水线取得 96.7% 准确率、0.968 的 F1-score 和 0.978 的 ROC-AUC。
正文 · 原文
Abstract:The arteriolar-to-venular diameter ratio (AVR) is a clinically motivated marker of retinal vascular health. We present a five-stage pipeline for AVR estimation and exploratory retinopathy screening from fundus images, with inspectable intermediate outputs at each stage. The pipeline combines vessel and artery-vein masks, optic-disc localization, vessel-calibre estimation using a minimal-chord boundary walk, and AVR computation from dominant arterial and venular branches. An explicit threshold-based rule produces the screening decision, enabling errors to be traced to individual processing stages. On an exploratory cohort of 30 HRF images, we report 96.7% accuracy, an F1-score of 0.968 and ROC-AUC of 0.978. Ablation studies and an exploratory comparison with VascX examine the effects of vessel pairing, calibre measurement and aggregation. Interpretation is limited by the small cohort, proxy disease labels and absence of independent reference calibre measurements. The pipeline provides a transparent framework for investigating retinal vascular biomarkers, with further validation needed on larger, clinically graded cohorts.
| Comments: | Accepted for presentation at the 23rd IEEE India Council International Conference (INDICON 2026) |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2611.14013 [cs.CV] |
| (or arXiv:2611.14013v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2611.14013 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Sharan Sharmeela Karuppannan [view email]
[v1]
Tue, 10 Nov 2026 12:51:27 UTC (3,293 KB)
来源：arXiv cs.AI · arxiv.org