Visible-Near Infrared Multispectral Imaging of the Rat Dorsal Skin Flap

Publisher Website
Google Scholar
Abstract
Visible-near infrared multispectral reflectance image sets were acquired from the dorsal surface of rats both before and after elevation of reversed McFarlane skin flaps. Raw images were dominated by uneven surface illumination and shadowing along with the variation associated with instrument response. These interfering features obscured variation associated with a change in tissue reflectance, which is related to the degree of flap perfusion. Logarithmic residual preprocessing followed by principal component analysis of multispectral images could clearly detect a difference in the optical properties between the base and distal section of the flap. The difference in the reflectance properties correlates with the varying degree of tissue perfusion. Principal component analysis detected this optical difference between the well-perfused base of the skin flap and the compromised distal section of the flap immediately following surgery. The first visual signs of compromised tissue perfusion appeared only 6 or more hours after surgery. The results from this study indicate that the application of principal component analysis to discrete wavelength near infrared multispectral reflectance images of skin flaps can effectively distinguish reflectance changes related to the degree of tissue perfusion immediately following surgical elevation of the reversed McFarlane skin flap. © 1999 Society of Photo-Optical Instrumentation Engineers.