A study of vein patterns in the forearm of Thai population using short-wave infrared light through image processing techniques

Abstract

The cubital superficial vein patterns vary among individuals, with the most significant difference being the lack of the median cubital vein, which is frequently used for diagnosis and treatment. Lack of understanding of vein patterns can put patients at risk and cause harm. In the past, vein pattern classification was achieved through observation. This study aims to develop a system using near-infrared imaging and image processing to divide the image into six zones and calculate the total area difference as a classification criterion. The study involved 150 Thai volunteers, and after excluding inappropriate images, the images were classified into 138 right arms and 112 left arms. The results of this research showed that there were four types of vein patterns, and pattern 3 was the most common (83 arms, 33.2%). The patterns of right and left arms were different, with pattern 1 being more common in the left arm than in the right arm. Among 103 subjects who had their vein patterns observed in both arms, 37 had the same pattern. The accuracy of the proposed method was compared to the accuracy of the previously studied method. From the results of the vein pattern classification from the developed system compared to the venous pattern classification by experts, it was found that the error was 1.4%, which was an acceptable value. Hence, it can be concluded that the proposed method can accurately characterize the vein pattern.