TY - JOUR
T1 - A CAD system for cerebral glioma based on texture features in DT-MR images
AU - De Nunzio, G.
AU - Pastore, G.
AU - Donativi, M.
AU - Castellano, A.
AU - Falini, A.
PY - 2011/8/21
Y1 - 2011/8/21
N2 - Tumor cells in cerebral glioma invade the surrounding tissues preferentially along white-matter tracts, spreading beyond the abnormal area seen on conventional MR images. Diffusion Tensor Imaging can reveal large peritumoral abnormalities in gliomas, which are not apparent on MRI. Our aim was to characterize pathological vs. healthy tissue in DTI datasets by 3D statistical Texture Analysis, developing an automatic segmentation technique (CAD, Computer Assisted Detection) for cerebral glioma based on a supervised classifier (an artificial neural network). A Matlab GUI (Graphical User Interface) was created to help the physician in the assisted diagnosis process and to optimize interactivity with the segmentation system, especially for patient follow-up during chemotherapy, and for preoperative assessment of tumor extension. Preliminary tissue classification results were obtained for the p map (the calculated area under the ROC curve, AUC, was 0.96) and the FA map (AUC=0.98). Test images were automatically segmented by tissue classification; manual and automatic segmentations were compared, showing good concordance.
AB - Tumor cells in cerebral glioma invade the surrounding tissues preferentially along white-matter tracts, spreading beyond the abnormal area seen on conventional MR images. Diffusion Tensor Imaging can reveal large peritumoral abnormalities in gliomas, which are not apparent on MRI. Our aim was to characterize pathological vs. healthy tissue in DTI datasets by 3D statistical Texture Analysis, developing an automatic segmentation technique (CAD, Computer Assisted Detection) for cerebral glioma based on a supervised classifier (an artificial neural network). A Matlab GUI (Graphical User Interface) was created to help the physician in the assisted diagnosis process and to optimize interactivity with the segmentation system, especially for patient follow-up during chemotherapy, and for preoperative assessment of tumor extension. Preliminary tissue classification results were obtained for the p map (the calculated area under the ROC curve, AUC, was 0.96) and the FA map (AUC=0.98). Test images were automatically segmented by tissue classification; manual and automatic segmentations were compared, showing good concordance.
KW - CAD
KW - Glioma
KW - Neural networks
KW - Texture features
UR - http://www.scopus.com/inward/record.url?scp=79960834644&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=79960834644&partnerID=8YFLogxK
U2 - 10.1016/j.nima.2010.12.086
DO - 10.1016/j.nima.2010.12.086
M3 - Article
AN - SCOPUS:79960834644
SN - 0168-9002
VL - 648
JO - Nuclear Instruments and Methods in Physics Research, Section A: Accelerators, Spectrometers, Detectors and Associated Equipment
JF - Nuclear Instruments and Methods in Physics Research, Section A: Accelerators, Spectrometers, Detectors and Associated Equipment
IS - SUPPL. 1
ER -