Abstract

Skin cancer full resection implies an evaluation of safety margins around the visible tumour. For melanomas such margins are proportional to tumour’s thickness also known as “Breslow Index”. In order to see if Diffuse Reflectance Spectroscopy (DRS) could be used to non-invasively evaluate Breslow Index, an in vitro study as well as numerical simulations were performed. Bilayered phantoms were made : a lower layer mimicking dermis underneath an absorbing layer mimicking a melanoma. Five groups of phantoms each having a specific top layer's thickness were made : 2, 3, 4, 5 or 6 mm. For wavelengths longer than 600 nm, Diffuse Reflectance spectra were significantly different (p<0.05) for each thickness at every Collecting to Excitation Fibre Separations (CEFS) : 271, 536, 834, 1076 and 1341 µm. Monte Carlo simulations were performed to check if DRS could detect smaller (i.e. 0.5 mm) thickness variations. Both experimental and numerical results showed the DR signal intensity linearly (R2>0.9) decreases as CEFS increases. The thicker the melanic layer was the smaller the slope (absolute value) was. These in vitro results will help setting up a clinical trial to non invasively evaluate Breslow Index : the bandwidth should be the NIR one (wavelengths longer than 600 nm) and CEFS should be shorter than 1 mm. Calibration will have to be made in order to relate slope to Breslow Index.

© 2007 SPIE