Optical Characterization and Multi-modality, Multi-scale Modeling of Human Skin Applied to Cancer Diagnosis.

Optical Characterization and Multi-modality, Multi-scale Modeling of Human Skin Applied to Cancer Diagnosis.

Skin carcinomas are the most commonly diagnosed cancers in fair-skinned populations, for example in France, Western Europe, and North America in particular. The OpticSkin project will build and make available to the general public and the scientific and medical community a histological and optical spectroscopic database of healthy, precancerous, and cancerous human skin in terms of absorption, elastic and inelastic scattering (Raman), steady-state and time-resolved autofluorescence, and polarization. The aim is to identify spectroscopic signatures that will be useful for diagnosis.

The Priority Research Program and Equipment for Light-Matter (PEPR LUMA) has granted its support to the OpitcSkin project, which aims to provide diagnostic information that complements that currently provided by histology for the diagnosis of skin carcinomas, the most common cancers among fair-skinned populations. In recent years, several imaging and optical spectroscopy modalities have been evaluated in vivo in clinical settings to quantify the real-time diagnostic assistance they provide to clinicians performing surgical resection of skin carcinomas. Imaging methods such as confocal reflectance microscopy, optical coherence tomography, and nonlinear optical microscopy provide morphological information that improves diagnostic accuracy and reduces the risk of recurrence by allowing immediate verification after surgery that the tumor has been completely removed.

Spectroscopic methods, including Raman spectroscopy and autofluorescence, which are also applied in vivo, provide additional structural and functional information, for example on metabolism, further improving diagnostic accuracy. In all cases, studies have shown that combining multiple optical imaging and/or spectroscopy modalities, as well as data analysis and/or machine learning methods, offers better diagnostic accuracy than each modality taken individually.

Inclusion Criteria:

• patients with skin carcinomas and actinic keratoses

Exclusion Criteria:

• minor patient