Most treatments and diagnoses are designed without properly considering highly pigmented skin (e.g. dark skin). For these skin types, the high melanin absorption in the epidermis, which is a thin skin layer, can cause overheating, and eventually cause skin damage. Our research focuses on understanding how the skin parameters (physiological and optical properties) can influence light-based treatments (photobiomodulation, photodynamic therapy, etc.) and optical diagnosis (e.g. pulse oximetry, optical fiber measurements), considering multiple skin phototypes. This kind of research can ultimately lead to the development of reformulated ways to safely light in health sciences, for all skin types.
Finding patterns on neoplastic lesions (like skin cancer and cervix lesions) can help in the development of new machine learning and artificial intelligence techniques for image classification, detection, and analysis. Exploring the features of different lesions through these techniques can bring novelties to the way we diagnose or interpret these lesions clinically. Our research focuses on digital image processing and artificial intelligence techniques to find patterns between different lesions. See our Python libraries, with the code already developed:
ImFun - Imaging Functions Library: github.com/MarlonGarcia/imfun
ImPro - Image Processing Functions: github.com/MarlonGarcia/imfun
Attacking White Blood Cells: github.com/MarlonGarcia/attacking-white-blood-cells
The way light is distributed into the tissues is highly important both for light-based treatments and for optical diagnosis. This light-biological tissue interaction is mainly governed by the optical properties of these tissues, which makes it essential to have an easy-to-use way of measuring the optical properties of tissues before being treated. This project focuses on the use of spatially-resolved diffuse reflectance to estimate these optical properties in vivo, a method that can be used in a clinical scenario.