Our approach combines spontaneous Raman spectroscopy with powerful 1-D convolutional neural network machine learning  classification to accurately diagnose unique metabolic subtypes of triple negative breast cancer.