Separable spectral unmixing based on the learning of periodic absorbance changes: Application to functional brain mapping using RGB imaging

Separable spectral unmixing designates techniques that allow to decompose spectra as a linear or non-linear combination of spectra of the targets (endmembers) collected. These techniques allow quantitative measurements but several drawbacks limit its use with standard optical devices like RGB cameras. We propose a new method for estimating endmembers and their proportion without calibration of the acquisition device with the analysis of periodic events in the signal. We evaluated the performances of the method for identifying functional brain areas during neurosurgery using RGB imaging. Results were consistent with clinical gold standards. This work can allow a widespread use of spectral imaging in the industrial or medical field.