ABSTRACT
Objective To assess the interobserver variations in delineating the planning target volume (PTV) and organs at risk (OAR) using different contouring methods during intensity-modulated radiation therapy (IMRT) for nasopharyngeal carcinoma (NPC),aiming to provide references for the quality control of multi-center clinical trials.Methods The PTV and OAR of CT image of 1 NPC patient manually delineated by 10 physicians from 8 different radiation centers were defined as the "manual contour group",and the OAR auto-contoured using the ABAS software and modified by the physicians were defined as the "auto+manual contour group".The maximum/minimum ratio (MMR) of the PTV and OAR volumes,and the coefficient of variation (CV) for different delineated contours were comparatively evaluated.Results Large variation was observed in the PTV and OAR volumes in the manual contour group.The MMR and CV of the PTV were 1.72-3.41 and 0.16-0.39,with the most significant variation in the PTVnd (MMR=3.41 and CV =0.39 for the PTVnd-L).The MMR and CV of the manually contoured OAR were 1.30-7.89 and 0.07-0.67.The MMR of the temporal lobe,spinal cord,temporomandibular joint,optic nerve and pituitary gland exceeded 2.0.Compared with the manual contour group,the average contouring time in the auto+ manual group was shortened by 68％ and the interobserver variation of the OAR volume was reduced with an MMR of 1.04-2.44 and CV of 0.01-0.37.Conclusions Large variation may occur in the PTV and OAR contours during IMRT plans for NPC delineated by different clinicians from multiple medical centers.Auto-contouring+ manually modification can reduce the interobserver variation of OAR delineation,whereas the variation in the delineation of small organs remains above 1.5 times.The consistency of the PTV and OAR delineation and the possible impact upon clinical outcomes should be reviewed and evaluated in multi-center clinical trials.
ABSTRACT
OBJECTIVE@#To establish a digital subtraction angiography (DSA) information management and image analysis system to realize scientific management of DSA image information and efficient processing of image data.@*METHODS@#Based on Java Web under Windows 7 environment, a dynamic Browser/Server mode system was constructed by JSP and Servlet on the network. Eclipse and MySQL were used as development tool and database development platform. Tomcat network information service was used as application server. Matlab codes were embedded to analyze DSA image.@*RESULTS@#The system consists of five modules:image information management, image processing, image analysis, advanced retrieval and clinical data management. It may complete such process as storing, deleting, saving, analyzing of DSA image and basic information of patients.@*CONCLUSIONS@#The main interface of the system is user-friendly and easy to operate. The system will be helpful to the clinical, teaching and scientific research work related to DSA.
Subject(s)
Humans , Angiography, Digital Subtraction , Databases, Factual , Indonesia , Information Management , Internet , Software , User-Computer InterfaceABSTRACT
Objective@#To assess the interobserver variations in delineating the planning target volume (PTV) and organs at risk (OAR) using different contouring methods during intensity-modulated radiation therapy (IMRT) for nasopharyngeal carcinoma (NPC), aiming to provide references for the quality control of multi-center clinical trials.@*Methods@#The PTV and OAR of CT image of 1 NPC patient manually delineated by 10 physicians from 8 different radiation centers were defined as the " manual contour group" , and the OAR auto-contoured using the ABAS software and modified by the physicians were defined as the " auto+ manual contour group" . The maximum/minimum ratio (MMR) of the PTV and OAR volumes, and the coefficient of variation (CV) for different delineated contours were comparatively evaluated.@*Results@#Large variation was observed in the PTV and OAR volumes in the manual contour group. The MMR and CV of the PTV were 1.72-3.41 and 0.16-0.39, with the most significant variation in the PTVnd (MMR=3.41 and CV=0.39 for the PTVnd-L). The MMR and CV of the manually contoured OAR were 1.30-7.89 and 0.07-0.67. The MMR of the temporal lobe, spinal cord, temporomandibular joint, optic nerve and pituitary gland exceeded 2.0. Compared with the manual contour group, the average contouring time in the auto+ manual group was shortened by 68% and the interobserver variation of the OAR volume was reduced with an MMR of 1.04-2.44 and CV of 0.01-0.37.@*Conclusions@#Large variation may occur in the PTV and OAR contours during IMRT plans for NPC delineated by different clinicians from multiple medical centers. Auto-contouring+ manually modification can reduce the interobserver variation of OAR delineation, whereas the variation in the delineation of small organs remains above 1.5 times. The consistency of the PTV and OAR delineation and the possible impact upon clinical outcomes should be reviewed and evaluated in multi-center clinical trials.