Abstract

Objective: The objective of this study was two folds: firstly, we would like to investigate the efficiency of bimodal spectroscopic technique in characterization of hypertrophic scarring tissue deliberately created on a preclinical model (rabbit’s ear); on the other hand, we evaluate the inhibition effect of an anti-inflammatory medication (tacrolimus) on hypertrophic formation in scar by using our bimodal spectroscopic system.

Study design: This study was conducted on 20 New Zealand Rabbits receiving hypertrophic scarring treatment on their ears. Fluorescence and Diffuse Reflectance spectra were collected from each scar, amongst which some had received tacrolimus treatment. Features were extracted from corrected spectral data and analyzed to classify the scarring tissues into hypertrophic or non-hypertrophic. Diagnostic algorithms were developed with the use of k-NN classifier and validated by comparing to histological classification result with Leave-one- out cross validation.

Results and discussion: The accuracy of our bimodal spectroscopy method for detecting hypertrophic scarring scar tissue was good (sensibility: 90.84%, specificity: 94.44%). The features used for classification were mainly extracted from the spectra exited at 360, 410 and 420 nm. This indicates that the difference between the spectra acquired from hypertrophic and non-hypertrophic tissue may be due to the different intensity distribution of several fluorophores (collagen,elastin and NADH) excited in this range, or to the change in proportion of tissue layers (epidermis and dermis) explored by the CEFS in use.

© 2011 OSA/SPIE