Tehran University of Medical Sciences

Science Communicator Platform

Stay connected! Follow us on X network (Twitter):
Share this content! On (X network) By
Video Densitometry Technique in Computation of Blood Flow Velocity in Diabetic Retinopathy Publisher



Farzad R1 ; Rassoul A1 ; Saeed S1 ; Farshad N2 ; Riazi EM3
Authors
Show Affiliations
Authors Affiliations
  1. 1. Digital Signal Processing Research Lab., Department of Electrical and Computer Engineering, Isfahan University of Technology, Isfahan, 84156-83111, Iran
  2. 2. Department of Electrical and Computer Engineering, University of Isfahan, Isfahan, 81746-73441, Iran
  3. 3. Department of Ophthalmology, Eye Research Center, Farabi Eye Hospital, Tehran University of Medical Sciences, Tehran, 13366-16351, Iran

Source: Journal of Medical Imaging and Health Informatics Published:2016


Abstract

Variations in the retinal vessels blood flow velocity represent one of the primary symptoms of diabetic retinopathy (DR). In this paper, we propose an algorithm to compute arterial blood flow velocity using florescent angiography video clips. The first step of this algorithm is to extract retinal vessels using Statistically Local Threshold (STL). After sharpening the arteries in the images, we can separate them from veins, and the main arteries are pruned using morphological methods. After that, some points are selected on the main arteries, considering their order on the vessel. In the next stage, having the coordinates of the selected points from the main arteries, we process angiographic videos to obtain intensity versus time curves. Then, relative intensity versus time is used instead of intensity versus time to reject attenuation artifact. Finally, intensity versus time curves are compared for consecutive points to obtain the cross time (CR) of dye. Based on the CR between points, our next step would be to compute on-vasculature distance (OVD) between pixels. And the blood flow velocity is calculated in vessels by CR and OVD. The proposed algorithm was tested on 16 florescent angiography video clips. The results indicate that blood flow in the main arteries has a range of 5.2 to 12.2 mm/sec in mild non-proliferative diabetic retinopathy (NPDR) patients, 1.1 to 8.5 mm/sec in moderate NPDR patients, 0.9 to 1.3 mm/sec in proliferative DR patients and 4.5 to 4.8 mm/sec in non-diabetic cases. Our results show that the blood flow velocity in patients with mild NPDR is obviously higher than healthy subjects. The algorithm was implemented in MATLAB software. © 2016 American Scientific Publishers All rights reserved.
Related Docs
1. Extraction of Vessel Structure in Thermal Images to Help Early Breast Cancer Detection, Computer Methods in Biomechanics and Biomedical Engineering: Imaging and Visualization (2020)