Abstract

Identifying vulnerable plaques at an early stage is critical to reducing patient mortality associated with cardiovascular diseases. Diffuse reflectance spectroscopy (DRS) can directly measure of absorption and scattering properties of tissue and can detect oxy-haemoglobin inside the plaque associated with intraplaque haemorrhage, which is a feature of vulnerable plaques. This study assesses the potential of using a diffuse reflectance spectroscopy combined with a machine learning algorithm to identify oxygenated haemoglobin in a chick embryo chorioallantoic membrane (CAM) assay model. A total of 88 diffuse reflectance spectra measurements were collected from five 12 to 14-day-old embryos. The first and second derivative of the reflectance spectra were calculated, followed by the use of partial least square-linear discriminant analysis (PLS-LDA) to identify oxygenated haemoglobin in chick embryo vessels. The model achieved a sensitivity and specificity of 96% and 72%, respectively, in differentiating arteries from veins (oxy-haemoglobin) using reflectance data. The sensitivity and specificity were 92% and 88% using the first derivative of reflectance data, and 100% and 92% using the second derivative of reflectance data in the wavelength range of 500-600 nm. Initial results indicate that derivative reflectance combined with multivariate analysis has advantages for detecting tissue oxygenated haemoglobin in CAM assay model. This approach shows promise as a way to identify and study the features of vulnerable plaques.

© 2023 SPIE