Abstract

Pulsed photothermal radiometry (PPTR) can be used for non-invasive depth profiling of skin vascular lesions (e.g., port wine stain birthmarks), aimed towards optimizing laser therapy on an individual patient basis. Optimal configuration of the experimental setup must be found and its performance characterized on samples with well defined structure, before introducing the technique into clinical practice. The aim of our study is to determine how sample structure and width of spectruml acquisition band affect the accuracy of measured depth profiles. We have constructed tissue phantoms composed of multiple layers of agar and of thin absorbing layers between the agar layers. Three phantoms had a single absorber layer at various depths between 100 and 500 μm, and one phantom had two absorber layers. In each sample we induced a non-homogeneous temperature profile with a 585 nm pulsed laser and acquired the resulting radiometric signal with a fast InSb infrared camera. We tested two configurations of the acquisition system, one using the customary 3-5 um spectruml band and one with a custom 4.5 pm cut-on filter. The laser-induced temperature depth profiles were reconstructed from measured PPTR signals using a custom algorithm and compared with sample structure as determined by histology and optical coherent tomography (OCT). PPTR determined temperature profiles correlate well with sample structure in all samples. Determination of the absorbing layer depth shows good repeatability with spatial resolution decreasing with depth. Spectruml filtering improved the accuracy of reconstructed profdes for shallow absorption layers (100-200 pm). PPTR technique enables reliable determination of structure in tissue phantoms with thin absorbing layers. Narrowing of the spectruml acquisition band (to 4.5 - 5.3 pm) improves reconstruction of objects near the surface.

© 2007 SPIE