Non-Invasive Analysis of Actinic Keratosis before and after Topical Treatment Using a Cold Stimulation and Near-Infrared Spectroscopy

Abstract

Background and objectives: The possible evolution of actinic keratoses (AKs) into invasive squamous cell carcinomas (SCC) makes their treatment and monitoring essential. AKs are typically monitored before and after treatment only through a visual analysis, lacking a quantitative measure to determine treatment effectiveness. Near-infrared spectroscopy (NIRS) is a non-invasive measure of the relative change of oxy-hemoglobin and deoxy-hemoglobin (O₂Hb and HHb) in tissues. The aim of our study is to determine if a time and frequency analysis of the NIRS signals acquired from the skin lesion before and after a topical treatment can highlight quantitative differences between the AK skin lesion area. Materials and Methods: The NIRS signals were acquired from the skin lesions of twenty-two patients, with the same acquisition protocol: baseline signals, application of an ice pack near the lesion, removal of ice pack and acquisition of vascular recovery. We calculated 18 features from the NIRS signals, and we applied multivariate analysis of variance (MANOVA) to compare differences between the NIRS signals acquired before and after the therapy. Results: The MANOVA showed that the features computed on the NIRS signals before and after treatment could be considered as two statistically separate groups, after the ice pack removal. Conclusions: Overall, the NIRS technique with the cold stimulation may be useful to support non-invasive and quantitative lesion analysis and regression after a treatment. The results provide a baseline from which to further study skin lesions and the effects of various treatments.

Keywords: actinic keratosis; field cancerization; near-infrared spectroscopy; signal processing.

Figure 1

The position of the probes in the actinic keratosis (AK) lesions (A) and healthy skin (B).

Figure 2

First row: example of AK before (A) and after the treatment (B), showing the typical AK features with erythema, telangiectasias, and slight hyperkeratotic scales (A), that were both strongly reduced after therapy (B). Second and last row: example of oxy-hemoglobin and deoxy-hemoglobin (O₂Hb and HHb) signals acquired on the AK lesion before (C) and after the treatment (D). In panels (C,D), the HHb signals are blue and the O₂Hb signals are red. Conc: micro moral (µMol) concentration.

Figure 3

The first two canonical variables of the features of the HHb and O₂Hb signals as calculated by MANOVA.