Author: Lily Jo Bertemes, Careesa Billante, Ashley Cetnar, Maximilian Stephen Meineke, Runhe Tan 👨🔬
Affiliation: The James Cancer Center, The Ohio State University, The Ohio State University - James Cancer Hospital 🌍
Purpose: The response of healthy vasculature due to radiation treatments is an underexplored area of radiation oncology. Understanding the impact of radiation on veins could advance treatment optimization. This study focuses on how a vein can be imaged noninvasively and digitally processed.
Methods: To image vasculature, the VeinViewer Flex was used. This device uses near-infrared light to locate veins beneath the skin at depths of up to 10mm. The VeinViewer then projects veins as a silhouette on the skin, which can be exported as an image file for digital processing. We used this device to image regions on a pig consisting of the back, inner leg, and ear. The images were combined to create a larger field of view using photopea.com, an online , forming composite images for each of the three regions. A python program was developed for post-processing the image to reduce noise, removing smaller structures (such as hair), and enhancing contrast of veins. The edges of the veins were detected using the cv2 library, and resulting images were displayed with pyplot.
Results: Veins from the pig’s ear and inner leg were detected with the device, while ones on the back of the pig were not visible. When visible, veins with diameters from 3.5-0.85 mm could be analyzed in python. Other aspects of the skin such as small hairs and pores were also identified. Smaller veins <0.85 mm led to uncertainty as they lacked contrast with the background.
Conclusion: Veins were able to be detected in the pig’s ear and inner leg, but not the back. Future study could explore other regions and implementation for human study. The program was able to detect veins in images and can be improved to reduce objects interfering with the current edge detection algorithm.