Hyperspectral evaluation of tissue vasculature structure and oxygenation

Blood vessel network structure and function are important indicators of a disease. Hyperspectral imaging (HSI) is an optical imaging modality combining imaging and spectroscopy offering insight into tissue structure and biochemical processes.

We present a novel vasculature imaging method where HSI are analyzed by B-COSFIRE method for automatic segmentation of linear structures. Normal and diseased tissues were imaged, blood distribution maps extracted from the images and used to determine vessel networks by the automatic segmentation. We present the protocol for generation of such maps along with examples from real-life studies.

Distinct differences between the normal and diseased tissue were observed, showing a large difference in the oxygenation. Defects in vasculature, such as obstructed vessels, were clearly visible from the generated maps. The determined vessel networks could serve further vasculature analysis, e.g. vessel fractal number or branching.

We demonstrated that analysis of tissue HSI by B-COSFIRE enables visualization of blood vessel network along with the oxygenation information. The approach can be extended to treatment planning in oncology, where vascularization of tumors can play a vital role in predicting treatment response.