False color images of cervical tissue sections, malignant glioma rat brain sections and melanoma sections have been constructed from FTIR hyperspectral data. The images are directly compared with stained sections and shown to be useful in distinguishing cervical pathology, malignant glioma and class I human lymphocyte antigen (HLA) expression in melanoma. Infrared spectra extracted from the images are shown to be useful for determining the macromolecular changes that accompany disease. Techniques for building three dimensional images to determine the extent of tissue change are described.