Dan-Ming Li1, Xin-Hu Wu, Xi-Xu Zhu. 1. Department of Radiation Oncology, Nanjing General Hospital of Nanjing Military Command, Nanjing, Jiangsu, 210002, People' Republic of China.
Abstract
BACKGROUND & OBJECTIVE: Magnetic resonance imaging (MRI) can well distinguish soft tissue, but its usage in radiotherapy for brain tumors was limited by its image distortion and lack of electron density for dosage calculation. The registration of MRI with computed tomography (CT) could resolve this problem. This study was to investigate the registration accuracy of MRI-CT image, and compare the change of center position, clinical target volume (CTV) and volume of organs at risk (OARs). METHODS: Nine glioma patients were examined by MRI and CT after operation. The registration accuracy of MRI-CT was evaluated. The changes of CTV, volume of OARs, and center position were investigated by the method of volume and center of geometry (COG) respectively. COG was used to measure the distance between center positions on MRI-CT registration image and CT location image. The volume method was used to measure the percentage of the volume of registration area (VMRI-CT) to the total volume (VMRI+CT). RESULTS: The registration accuracy by the method of artificial landmarks was less than 1.5 mm, which reached the error requirement for brain tumors. No significant change of volume of OARs was found (P>0.05). Of the 9 patients, 8 had decreased CTV (by 13.85%-73.59%) on MRI-CT registration image and 1 had increased CTV (by 10.35%). The difference in mean CTV between the 2 groups was significant (P<0.05). The change of central position of CTV was the largest [(8.74+/-6.60) mm], those of 2 eyes were the next [(5.25+/-2.38) mm for the left eyeû (5.65+/-2.56) mm for the right eye], and that of the brain stem was the least [(1.83+/-1.06) mm]. CONCLUSIONS: By the method of artificial landmarks, MRI-CT has better registration accuracy. MRI-CT registration can reduce the uncertainty of CTV delineation in radiation treatment planning for glioma patients after operation.
BACKGROUND & OBJECTIVE: Magnetic resonance imaging (MRI) can well distinguish soft tissue, but its usage in radiotherapy for brain tumors was limited by its image distortion and lack of electron density for dosage calculation. The registration of MRI with computed tomography (CT) could resolve this problem. This study was to investigate the registration accuracy of MRI-CT image, and compare the change of center position, clinical target volume (CTV) and volume of organs at risk (OARs). METHODS: Nine gliomapatients were examined by MRI and CT after operation. The registration accuracy of MRI-CT was evaluated. The changes of CTV, volume of OARs, and center position were investigated by the method of volume and center of geometry (COG) respectively. COG was used to measure the distance between center positions on MRI-CT registration image and CT location image. The volume method was used to measure the percentage of the volume of registration area (VMRI-CT) to the total volume (VMRI+CT). RESULTS: The registration accuracy by the method of artificial landmarks was less than 1.5 mm, which reached the error requirement for brain tumors. No significant change of volume of OARs was found (P>0.05). Of the 9 patients, 8 had decreased CTV (by 13.85%-73.59%) on MRI-CT registration image and 1 had increased CTV (by 10.35%). The difference in mean CTV between the 2 groups was significant (P<0.05). The change of central position of CTV was the largest [(8.74+/-6.60) mm], those of 2 eyes were the next [(5.25+/-2.38) mm for the left eyeû (5.65+/-2.56) mm for the right eye], and that of the brain stem was the least [(1.83+/-1.06) mm]. CONCLUSIONS: By the method of artificial landmarks, MRI-CT has better registration accuracy. MRI-CT registration can reduce the uncertainty of CTV delineation in radiation treatment planning for gliomapatients after operation.