Malignant-lesion segmentation using 4D co-occurrence texture analysis applied to dynamic contrast-enhanced magnetic resonance breast image data Export

@article{citeulike:1140790, abstract = {To investigate the use of four-dimensional (4D) co-occurrence-based texture analysis to distinguish between nonmalignant and malignant tissues in dynamic contrast-enhanced (DCE) MR images.4D texture analysis was performed on DCE-MRI data sets of breast lesions. A model-free neural network-based classification system assigned each voxel a ?nonmalignant? or ?malignant? label based on the textural features. The classification results were compared via receiver operating characteristic (ROC) curve analysis with the manual lesion segmentation produced by two radiologists (observers 1 and 2).The mean sensitivity and specificity of the classifier agreed with the mean observer 2 performance when compared with segmentations by observer 1 for a 95\% confidence interval, using a two-sided t-test with ? = 0.05. The results show that an area under the ROC curve (Az) of 0.99948, 0.99867, and 0.99957 can be achieved by comparing the classifier vs. observer 1, classifier vs. union of both observers, and classifier vs. intersection of both observers, respectively.This study shows that a neural network classifier based on 4D texture analysis inputs can achieve a performance comparable to that achieved by human observers, and that further research in this area is warranted. J. Magn. Reson. Imaging 2007. {\copyright} 2007 Wiley-Liss, Inc.}, address = {Department of Electrical and Computer Engineering, Ohio State University, Columbus, Ohio, USA; Department of Biomedical Engineering, Ohio State University, Columbus, Ohio, USA; Department of Biomedical Informatics, Ohio State University, Columbus, Ohio, USA; Department of Radiology, Ohio State University, Columbus, Ohio, USA; Department of Electrical and Electronics Engineering, Bilkent University, Ankara, Turkey}, author = {Woods, Brent J. and Clymer, Bradley D. and Kurc, Tahsin and Heverhagen, Johannes T. and Stevens, Robert and Orsdemir, Adem and Bulan, Orhan and Knopp, Michael V.}, citeulike-article-id = {1140790}, citeulike-linkout-0 = {http://dx.doi.org/10.1002/jmri.20837}, citeulike-linkout-1 = {http://www3.interscience.wiley.com/cgi-bin/abstract/114111141/ABSTRACT}, doi = {10.1002/jmri.20837}, journal = {Journal of Magnetic Resonance Imaging}, keywords = {application, med\_app, texture}, number = {3}, pages = {495--501}, posted-at = {2008-11-13 04:00:41}, priority = {2}, title = {Malignant-lesion segmentation using 4D co-occurrence texture analysis applied to dynamic contrast-enhanced magnetic resonance breast image data}, url = {http://dx.doi.org/10.1002/jmri.20837}, volume = {25}, year = {2007} }

TY - JOUR ID - citeulike:1140790 L3 - citeulike-article-id:1140790 N2 - To investigate the use of four-dimensional (4D) co-occurrence-based texture analysis to distinguish between nonmalignant and malignant tissues in dynamic contrast-enhanced (DCE) MR images.4D texture analysis was performed on DCE-MRI data sets of breast lesions. A model-free neural network-based classification system assigned each voxel a ?nonmalignant? or ?malignant? label based on the textural features. The classification results were compared via receiver operating characteristic (ROC) curve analysis with the manual lesion segmentation produced by two radiologists (observers 1 and 2).The mean sensitivity and specificity of the classifier agreed with the mean observer 2 performance when compared with segmentations by observer 1 for a 95% confidence interval, using a two-sided t-test with ? = 0.05. The results show that an area under the ROC curve (Az) of 0.99948, 0.99867, and 0.99957 can be achieved by comparing the classifier vs. observer 1, classifier vs. union of both observers, and classifier vs. intersection of both observers, respectively.This study shows that a neural network classifier based on 4D texture analysis inputs can achieve a performance comparable to that achieved by human observers, and that further research in this area is warranted. J. Magn. Reson. Imaging 2007. © 2007 Wiley-Liss, Inc. IS - 3 JF - Journal of Magnetic Resonance Imaging AD - Department of Electrical and Computer Engineering, Ohio State University, Columbus, Ohio, USA; Department of Biomedical Engineering, Ohio State University, Columbus, Ohio, USA; Department of Biomedical Informatics, Ohio State University, Columbus, Ohio, USA; Department of Radiology, Ohio State University, Columbus, Ohio, USA; Department of Electrical and Electronics Engineering, Bilkent University, Ankara, Turkey EP - 501 TI - Malignant-lesion segmentation using 4D co-occurrence texture analysis applied to dynamic contrast-enhanced magnetic resonance breast image data VL - 25 SP - 495 KW - application KW - med_app KW - texture AU - Woods, Brent AU - Clymer, Bradley AU - Kurc, Tahsin AU - Heverhagen, Johannes AU - Stevens, Robert AU - Orsdemir, Adem AU - Bulan, Orhan AU - Knopp, Michael PY - 2007/// UR - http://dx.doi.org/10.1002/jmri.20837 ER -