Hybrid Monte Carlo and machine learning solution for quantitative assessment of light wave localization in turbid media (Conference Presentation)

Designing a bespoke optical diagnostics/treatment system requires selecting various technical aspects such as intensity, wavelength, polarization, coherence, wavefront shape, orbital angular momentum, and quantum entanglement of light. Classical and non-classical phenomena of optical radiation interacting with biological tissues must be considered. This talk emphasizes our recent development of a hybrid generalized algorithm combining Monte Carlo simulations with emerging Machine Learning paradigms. Designed to run on low-power processors, this advanced approach bridges the gap between forward and inverse computations. Extensive validation against standard modelling/experimental techniques will be provided. Practical applications include optimized light delivery for photodynamic melanoma treatment, multispectral polarized imaging, 3D rendering of human skin, and leveraging complex light/quantum entanglement in biomedical diagnostics.