MDCT of Pancreatic Adenocarcinoma: Optimal Imaging Phases and Multiplanar Reformatted Imaging Export

@article{citeulike:997734, abstract = {OBJECTIVE. The objective of our study was to evaluate the individual contributions of arterial, pancreatic parenchymal, and portal venous phase (PVP) images and the utility of coronal and sagittal multiplanar reformatted (MPR) images in the assessment of pancreatic adenocarcinoma using triple-phase MDCT. MATERIALS AND METHODS. Thirty-one patients with and 35 patients without pancreatic adenocarcinoma underwent triple-phase MDCT. Three radiologists independently attempted to detect pancreatic adenocarcinoma and assess local extension using the MDCT images in five sessions. The first three sessions involved sets of images obtained in arterial phase, pancreatic parenchymal phase, and PVP separately and respectively. In the fourth session, a combination of axial images from all phases was evaluated. During the fifth session, radiologists had access to coronal and sagittal MPR images together with the axial images obtained in all phases. Results were compared with surgical findings using receiver operating characteristic (ROC) analysis and kappa statistics. RESULTS. Regarding tumor detection, the image set composed of coronal and sagittal MPR images and of axial images obtained in all phases had a significantly higher value for the area under the ROC curve (Az, 0.98 {+/-} 0.01) than the other image sets and yielded the highest sensitivity (93.5\%). The sensitivity of the arterial phase image set (80.6\%) was significantly lower than that of all other image sets. Whereas the image set composed of coronal and sagittal MPR images and axial images obtained in all phases yielded the highest kappa values for all local extension factors evaluated, the image set composed of only arterial phase images yielded the lowest kappa values for almost all of the factors. CONCLUSION. A combination of pancreatic parenchymal phase and PVP imaging is necessary and efficient for the assessment of pancreatic adenocarcinoma. The addition of coronal and sagittal MPR images increased the performance of MDCT, especially in the evaluation of local extension. 10.2214/AJR.05.1031}, author = {Ichikawa, Tomoaki and Erturk, Sukru M. and Sou, Hironobu and Nakajima, Hiroto and Tsukamoto, Tatsuaki and Motosugi, Utarou and Araki, Tsutomu}, citeulike-article-id = {997734}, citeulike-linkout-0 = {http://dx.doi.org/10.2214/AJR.05.1031}, citeulike-linkout-1 = {http://www.ajronline.org/cgi/content/abstract/187/6/1513}, citeulike-linkout-2 = {http://view.ncbi.nlm.nih.gov/pubmed/17114545}, citeulike-linkout-3 = {http://www.hubmed.org/display.cgi?uids=17114545}, day = {1}, doi = {10.2214/AJR.05.1031}, journal = {Am. J. Roentgenol.}, keywords = {carcinoma, ct, pancreas, pancreatic, technique}, month = {December}, number = {6}, pages = {1513--1520}, posted-at = {2006-12-16 12:45:54}, priority = {5}, title = {MDCT of Pancreatic Adenocarcinoma: Optimal Imaging Phases and Multiplanar Reformatted Imaging}, url = {http://dx.doi.org/10.2214/AJR.05.1031}, volume = {187}, year = {2006} }

TY - JOUR ID - citeulike:997734 L3 - citeulike-article-id:997734 N2 - OBJECTIVE. The objective of our study was to evaluate the individual contributions of arterial, pancreatic parenchymal, and portal venous phase (PVP) images and the utility of coronal and sagittal multiplanar reformatted (MPR) images in the assessment of pancreatic adenocarcinoma using triple-phase MDCT. MATERIALS AND METHODS. Thirty-one patients with and 35 patients without pancreatic adenocarcinoma underwent triple-phase MDCT. Three radiologists independently attempted to detect pancreatic adenocarcinoma and assess local extension using the MDCT images in five sessions. The first three sessions involved sets of images obtained in arterial phase, pancreatic parenchymal phase, and PVP separately and respectively. In the fourth session, a combination of axial images from all phases was evaluated. During the fifth session, radiologists had access to coronal and sagittal MPR images together with the axial images obtained in all phases. Results were compared with surgical findings using receiver operating characteristic (ROC) analysis and kappa statistics. RESULTS. Regarding tumor detection, the image set composed of coronal and sagittal MPR images and of axial images obtained in all phases had a significantly higher value for the area under the ROC curve (Az, 0.98 {+/-} 0.01) than the other image sets and yielded the highest sensitivity (93.5%). The sensitivity of the arterial phase image set (80.6%) was significantly lower than that of all other image sets. Whereas the image set composed of coronal and sagittal MPR images and axial images obtained in all phases yielded the highest kappa values for all local extension factors evaluated, the image set composed of only arterial phase images yielded the lowest kappa values for almost all of the factors. CONCLUSION. A combination of pancreatic parenchymal phase and PVP imaging is necessary and efficient for the assessment of pancreatic adenocarcinoma. The addition of coronal and sagittal MPR images increased the performance of MDCT, especially in the evaluation of local extension. 10.2214/AJR.05.1031 IS - 6 JF - Am. J. Roentgenol. EP - 1520 TI - MDCT of Pancreatic Adenocarcinoma: Optimal Imaging Phases and Multiplanar Reformatted Imaging VL - 187 SP - 1513 KW - carcinoma KW - ct KW - pancreas KW - pancreatic KW - technique AU - Ichikawa, Tomoaki AU - Erturk, Sukru AU - Sou, Hironobu AU - Nakajima, Hiroto AU - Tsukamoto, Tatsuaki AU - Motosugi, Utarou AU - Araki, Tsutomu PY - 2006/12/01/ UR - http://dx.doi.org/10.2214/AJR.05.1031 ER -