RESUMO
Objective:To effectively quantify and evaluate the quality of different deformation registration algorithms, in order to enhance the possibility of implementing deformation registration in clinical practice.Methods:The Jacobian determinant mean (JDM) is proposed based on the Jacobian determinant (JD) of displacement vector field (DVF), and the Jacobian determinant error (DJDE) is introduced by incorporating the JD of the inverse DVF. The optical flow method (OF-DIR) and fast demons method with elastic regularization (FD-DIR) were tested on nasopharyngeal and lung cancer datasets. Finally, JDM and DJDE with the Jacobian determinant negative percentage (JDNP), inverse consistency error (ICE) and normalized mean square error (NMSE) were used to evaluate the registration algorithms and compare the differences evaluation indicators in different tumor images and different algorithms, and the receiver operating curve (ROC) was analyzed in evaluation.Results:In lung cancer, OF-DIR outperformed FD-DIR in terms of JDM, NMSE, DJDE and ICE, and the difference was statistically significant( z = -2.24, -4.84, t = 4.01, 6.54, P<0.05). In nasopharyngeal carcinoma, DJDE, ICE and NMSE of OF-DIR were superior to FD-DIR, and the difference was statistically significant ( t = 4.46, -7.49, z = -2.22, P<0.05), but there was no significant difference in JDM ( P>0.05). In lung cancer and nasopharyngeal carcinoma, JDNP of OF-DIR was worse than that of FD-DIR, and the difference was statistically significant ( z = -4.29, -4.02, P<0.01). In addition, DJDE is more specific and sensitive on ROC curve (AUC=0.77), and has different performance result for tumor images at different sites. Conclusions:The JDM and DJDE evaluation metrics proposed are effective for deformation registration algorithms. OF-DIR is suitable for both lung cancer and nasopharyngeal carcinoma, while the influence of organ motion on the registration effect should be considered when using FD-DIR.
RESUMO
Objective:To identify the anatomical landmarks with stable and consistent performance in both rigid and deformation registrations,and to investigate the feasibility that take them as the quantitative indexes for quantitative evaluation of multimodality image registration for pancreatic cancer in clinical practice. Methods:Twenty-five patients with pancreatic cancer were selected.Both the rigid and deformation registrations were performed between simulated CT and MRI T1-weighted images.Seven anatomical landmarks,which were left adrenal gland,portal vein,celiac trunk artery,superior mesenteric artery,lumbar vertebral body,inferior vena cava and abdominal aorta,were selected.The anatomical landmarks were marked on the simulated CT and the registered MRI,respectively.The distances between the geometric centers of each anatomical landmarks in two kinds of images were statistically analyzed.The Wilcoxon rank sum test was implemented to compare the differences between rigid and deformation registrations.In addition,the correlations between anatomical landmarks were analyzed as well. Results:The mean distances of centers for the seven anatomical markers under rigid registration ranged from 0.405 cm to 1.097 cm,while they ranged from 0.433 cm to 0.740 cm under deformation registration.There was no significant difference in the registration of anatomical markers between rigid and deformation registration,except the left adrenal gland.The registration differences of celiac trunk artery and abdominal aorta were of most stable in 25 patients.Correlation study showed that all anatomical markers except abdominal aorta and inferior vena cava were independent and significant. Conclusion:It is feasible to quantitatively evaluate the registration quality of multimodality image registration for pancreatic cancer by using anatomical landmarks.It is recommended to use celiac trunk artery and abdominal aorta as the anatomical markers in clinical procedure.They are both stable in rigid and deformation registration,and the correlation between them is low,with means they can be used as independent evaluation criteria.
RESUMO
Objective To investigate the changes of accumulated dose in target area and organs at risk (OARs) for radiotherapy of left breast cancer by deformable and rigid image registration.Methods A total of 16 left breast cancer patients treated with 6 MV X-ray IMRT were analyzed retrospectively.All targets included the lymph node drainage area and the chest wall.All patients underwent simulation of the primary positioning and repositioning to obtain CT images.Primary and secondary treatment plans were developed using primary positioning CT (CT1) and repositioning CT (CT2),denoted as Plan1 and Plan2 respectively.The dose distribution of Plan2 was mapped to CT1 with rigid and deformable registration from CT2 to CT1 and then added to the dose distribution of Planl to obtain Plan-rigid and Plan-deform,respectively.The dosimetric differences between targets and the OARs of the four plans were compared.Results The CTV volume on CT2 was reduced by 6.64％ from that on CT1.The homogeneity index (HI)increased by 23.05％ after deformation-based accumulation.The Dice similarity coefficients (DSCs) of the heart,left lung and right lung were lower than those before deformable registration (0.94±0.01 vs.0.89± 0.05,0.96±0.01 vs.0.91±0.03,and 0.96±0.01 vs.0.92±0.03,respectively),and the differences were statistically significant (Z =-3.208,-3.533,-3.535,P ＜ 0.05).There were no significant differences in dose-volume indices of heart and left lung between Plan1 with other plans (P＞0.05),while the dose-volume indices in Plan-rigid were higher than that in Plan-deform.Conclusions Rigid registration is recommended in patients undergoing radical resection of left breast cancer with little change in the volume and dose-volume index of the target area and organs at risk.The dose-volume index of the initial intensity modulation plan can basically reflect the dose-volume statistics of both lungs and heart.
RESUMO
Objective@#To investigate the changes of accumulated dose in target area and organs at risk (OARs) for radiotherapy of left breast cancer by deformable and rigid image registration.@*Methods@#A total of 16 left breast cancer patients treated with 6 MV X-ray IMRT were analyzed retrospectively. All targets included the lymph node drainage area and the chest wall. All patients underwent simulation of the primary positioning and repositioning to obtain CT images. Primary and secondary treatment plans were developed using primary positioning CT (CT1) and repositioning CT (CT2), denoted as Plan1 and Plan2 respectively. The dose distribution of Plan2 was mapped to CT1 with rigid and deformable registration from CT2 to CT1 and then added to the dose distribution of Plan1 to obtain Plan-rigid and Plan-deform, respectively. The dosimetric differences between targets and the OARs of the four plans were compared.@*Results@#The CTV volume on CT2 was reduced by 6.64% from that on CT1. The homogeneity index (HI) increased by 23.05% after deformation-based accumulation. The Dice similarity coefficients (DSCs) of the heart, left lung and right lung were lower than those before deformable registration (0.94±0.01 vs. 0.89±0.05, 0.96±0.01 vs. 0.91±0.03, and 0.96±0.01 vs. 0.92±0.03, respectively), and the differences were statistically significant (Z=-3.208, -3.533, -3.535, P<0.05). There were no significant differences in dose-volume indices of heart and left lung between Plan1 with other plans(P>0.05), while the dose-volume indices in Plan-rigid were higher than that in Plan-deform.@*Conclusions@#Rigid registration is recommended in patients undergoing radical resection of left breast cancer with little change in the volume and dose-volume index of the target area and organs at risk. The dose-volume index of the initial intensity modulation plan can basically reflect the dose-volume statistics of both lungs and heart.