Detection of the brain tumor margins is one of the most significant problems in neurosurgery. Several mass spectrometry-based approaches have been proposed recently for tumor boundary detection. One of them, spray from tissue does not require sample preparation but needs special algorithms for analysis of its spectra. Here we proposed the feature selection algorithm designed for analysis of spray-from-tissue data.
Keywords: Direct mass spectrometry; ambient ionization; feature selection; neurosurgery; spray from tissue.