Statistical Analysis of Multispectral NIR Images for Vein Detection

Vein localization is crucial for successful cannulation, but it can be challenging due to factors such as skin colour and Body Mass Index (BMI). While near-infrared (NIR) imaging is established for improving vein visibility, the optimal wavelength for patients with varying skin colours and BMIs has not been thoroughly studied. Thus, it is important to explore optimal wavelength selection to address these influencing factors. In this study, we developed and tested a method for enhancing vein contrast using Multispectral NIR imaging and evaluated the impact of skin colour and BMI on vein visibility in a sample of 61 subjects. Our study found that the optimal wavelength for vein visibility in the NIR region was 864 nm, and the proposed method effectively enhanced vein contrast. We also found that skin colour and BMI influenced vein visibility, with fair skin and lower BMI associated with better vein visibility. Our study demonstrates the potential of NIR imaging for enhancing vein visibility, and also emphasizes in considering skin colour and BMI as significant factors. However, there were some limitations of our study, as the majority of the age group consisted of the young population and it was not inclusive of babies and elderly people for whom venous access is difficult.