This paper presents an MRI feature-space image-analysis method and its
application to brain tumor studies. The proposed method generates a t
ransformed feature space in which the normal tissues (white matter, gr
ay matter, and CSF) become orthonormal. As such, the method is expecte
d to have site-to-site and patient-to-patient consistency, and is usef
ul for identification of tissue types, segmentation of tissues, and qu
antitative measurements on tissues, The steps of the work accomplished
are as follows: (1) Four T-2-weighted and two T-1-weighted images (be
fore and after injection of gadolinium) were acquired for 10 tumor pat
ients. (2) Images were analyzed by an image analyst according to the p
roposed algorithm, (3) Biopsy samples were extracted from each patient
and were subsequently analyzed by the pathology laboratory. (4) Image
-analysis results were compared with the biopsy results. Pre- and post
surgery feature spaces were also compared, The proposed method made it
possible to visualize the MRI feature space and to segment the image.
In all cases, the operators were able to find clusters for normal and
abnormal tissues. Also, clusters for different zones of the tumor wer
e found. The method successfully segmented the image into normal tissu
es (white matter, gray matter, and CSF) and different zones of the les
ion (tumor, cyst, edema, radiation necrosis, necrotic core, and infilt
rated tumor). The results agreed with those obtained from the biopsy s
amples. Comparison of pre- with postsurgery and radiation feature spac
es illustrated that the original solid tumor was not present in the se
cond study, but a new tissue component appeared in a different locatio
n of the feature space. This tissue could be radiation necrosis genera
ted as a result of radiation.