Robust demarcation of basal cell carcinoma by dependent component analysis-based segmentation of multi-spectral fluorescence images.

This study was designed to demonstrate robust performance of the novel dependent component analysis (DCA)-based approach to demarcation of the basal cell carcinoma (BCC) through unsupervised decomposition of the red-green-blue (RGB) fluorescent image of the BCC. Robustness to intensity fluctuation is due to the scale invariance property of DCA algorithms, which exploit spectral and spatial diversities between the BCC and the surrounding tissue. Used filtering-based DCA approach represents an extension of the independent component analysis (ICA) and is necessary in order to account for statistical dependence that is induced by spectral similarity between the BCC and surrounding tissue. This generates weak edges what represents a challenge for other segmentation methods as well. By comparative performance analysis with state-of-the-art image segmentation methods such as active contours (level set), K-means clustering, non-negative matrix factorization, ICA and ratio imaging we experimentally demonstrate good performance of DCA-based BCC demarcation in two demanding scenarios where intensity of the fluorescent image has been varied almost two orders of magnitude.

Fractionated illumination improves the outcome in the treatment of precancerous lesions with photodynamic therapy.

Photodynamic therapy (PDT) is a noninvasive therapy for non-melanoma skin cancer. The aim of this study was comparison of efficacy between fractioned versus single dose illumination in photodynamic therapy (PDT) of actinic keratosis (AK) and Bowen's disease (BD). Fifty-one patients (36 AK and 15 BD) were treated with PDT They were randomly arranged in two treatment groups. Group one included 26 patients (20 AK and 6 BD) that, after five hours of incubation with 20% 5-ALA, were treated with a single illumination of 100 Jcm(-2) at fluence rate of 30 mWcm(-2). Group two included 25 patients (16 AK and 9 BD) that, after 16 hours of incubation with 20% 5-ALA, were treated with two light fractions (50 plus 50 Jcm(-2)) at same fluence rate with dark interval of two hours between fractions. Twenty-four weeks later, a treated area was incubated for four hours again with 5-ALA in order to detect occult areas of abnormal skin with possible remaining tumor tissue. In case of fluorescence, histological examination was performed. In the group one, fluorescence at the end of the session was absent in 19 (73%) or very weak in 7 (27%). Residual tumor was found in 15 (75%) AK and in 4 (66.6%) BD. In the group two, fluorescence at the end of second session was more intense; in one patient (4%) was absent, very weak in 5 (20%) and weak in 19 (76%) of patients. In this group histology revealed remaining tumor tissue in only 2 (12.5%) AK and 2 (22.2%) BD. Among the patients in the first group, the remaining tumor tissue was significantly bigger (p=0.005). The treatment response with clearing of tumor tissue was significantly higher in fractionated illumination than in a single dose illumination group. Fractionated illumination scheme with 16 hours of incubation separated by two hours dark interval significantly improves the therapeutic outcome in tumor eradication.

Visualization of basal cell carcinoma by fluorescence diagnosis and independent component analysis.

Photodynamic detection (PDD) of skin tumours is based on the visualization of a fluorophores, with the ability to accumulate in tumour tissue, by the use of fluorescence imaging. Of particular importance is the application of δ-5-aminolaevulinic acid (ALA) that, through the process of biosynthesis causes formation of the protoporphyrin IX (PpIX). The PpIX has the ability of selective fluorescence after basal cell carcinoma (BCC) has been treated with ALA. Higher concentration of PpIX in tumour tissue compared to surrounding normal skin is the basis for PDD. Our contribution in this preliminary study is application of the independent component analysis (ICA) to extract the BCC spatial map, by processing fluorescent RGB image acquired under excitation with 405nm light. Comparative performance analysis with other two widely used image processing methods: ratio imaging and optimal threshold based imaging, reveals that ICA produces BCC spatial map that is most consistent in term of diagnostic quality by both visual assessment and calculation of the BCC demarcation line. We believe this represents a solid basis for the design of a compact and low-cost multi-spectral fluorescence imaging system, capable for real time calculation of the skin tumour demarcation.