Red, green, and blue gray-value shift-based approach to whole-field imaging for tissue diagnostics

Abstract

Identification of abnormal pathology in situ remains one of the challenges of medicine. The interpretation of tissue conditions relies mainly on optical assessment, which can be difficult due to inadequate visual differences or poor color delineation. We propose a methodology to identify regions of abnormal tissue in a targeted area based on red, green, blue (RGB) shift analysis employing a simple CCD color camera and light-emitting diode illumination in a whole-field-imaging scheme. The concept involves analysis of RGB components in an image with respect to a reference set of RGB values under different illumination wavelengths. The magnitude of the gray value shift is estimated by calculating the Euclidean distance between their normalized RGB coordinates. The shift values obtained using these concepts are thereafter used to construct pseudo-colored images with high contrast, enabling easy identification of abnormal areas in the tissue. Images processed from experiments conducted with excised Wistar rat colon sample (lightly doped with Alexafluor 488) and with simulated tumor (cancer cell pellet placed on colon) showed clear localization of tumor region. This proposed approach and methodology is expected to find potential applications for the in vivo diagnosis of disease.