Application Of Dual - Source CT Dual Energy Imaging And Iterative Reconstruction In Pulmonary Embolism

Objective: To explore the value of application of dual-energy CT sinogram-affirmed iterative Reconstruction and non-Linear blending technique in improving image quality of CT Pulmonary Artery Angiography.Methods: 30 patients were examined by Germany Siemens dual-source CT in dual energy mode. The data were reconstructed by filtered back projection mode and 100 k V, 140 k V images and linear integration data whose mixing ratio was 3:2 were obtained. The data were reconstructed by 1-5 levels of sinogram-affirmed iterative reconstruction mode and 100 k V, 140 k V images and linear integration data whose mixing ratio was 3:2 were obtained. 100 k V,140 k V images of SAFIRE1-5 levels were reconstructed by contrast optimal software in dual-energy mode to generate a non-linear fusion images. Main pulmonary artery was used as a region of interest(ROI) and was measured in the images obtained by different reconstruction methods.At the same time,Vertical spinal muscular around the spine were measured as the reference region. CT values and the noise were measured. ROIM and ROIA respectively represented CT values of main pulmonary artery and muscular. The noise of main pulmonary artery was represented by SD. The signal to noise ratio(SNR)was ROIA/SD; the relative signal to noise ratio(CNR) was ROIA-ROIM/SD.The noise, SNR and CNR of linear integration images reconstructed by FBP mode and SAFIRE1-5 levels mode were analyzed by statistics software. The best reconstructed technique was screened. The noise, SNR and CNR of linear integration images reconstructed by FBP mode, best SAFIRE level mode and non-linear integration images reconstructed by best SAFIRE level mode were analyzed by statistics software.Meanwhile images with embolus in the pulmonary artery were subjectively scored. Image’s definition and image quality which included linear integration images reconstructed by FBP mode,best SAFIRE level mode and non-linear integration images reconstructed by best SAFIRE level mode were subjectively scored. The scores were analyzed by statistics software.Result: The noise of Safire reconstructed images was lower than it of FBP reconstruction images.SNR and CNR of Safire reconstructed images was higher than them of FBP reconstruction images. With increasing level of reconstruction of SAFIRE technique, noise decreased gradually and SNR and CNR increased gradually. The noise, SNR and CNR of images reconstructed by FBP and SAFIRE1-5 technique were statistically significantwith each other.(p<0.05) The noise of SAFIRE5 non-linear fusion images was lowest and CNR of them were highest. The noise of FBP linear fusion images was highest and SNR, CNR of them were the lowest. The noise, SNR and CNR between SAFIRE5 non-linear fusion images and FBP linear fusion images were statistically significant(p <0.05). And the noise, SNR and CNR between SAFIRE5 linear fusion images and FBP linear fusion images were statistically significant.(p <0.05) But the noise, SNR and CNR between SAFIRE5 linear fusion images and SAFIRE5 non-linear fusion images were not statistically significant.(p >0.05) The number of images with embolus in the pulmonary artery was 24 cases. The score of SAFIRE5 non-linear fusion images was the highest in evaluating their definition and image quality. And the score of FBP linear fusion image was the lowest in evaluating their definition and image quality. The scores of images in evaluating their definition and image quality between SAFIRE5 non-linear fusion images and FBP linear fusion images were statistically significant.(p <0.05) The scores of images in evaluating their definition and image quality between SAFIRE5 linear fusion images and FBP linear fusion images were statistically significant.(p <0.05) The scores of images in evaluating their definition between SAFIRE5 linear fusion images and SAFIRE5 non-linear fusion images were statistically significant.(p <0.05)But The scores in evaluating image quality between SAFIRE5 linear fusionimages and SAFIRE5 non-linear fusion images were not statistically significant.(p >0.05)Conclusion: SAFIRE iterative reconstruction technique could significantly reduce image noise, improve signal-to-noise ratio comparing with filter back projection reconstruction technique(FBP). SAFIRE5 reconstruction has maximum value to reduce image noise, improve signal-to-noise ratio.Combined application of SAFIRE iterative reconstruction and non-linear fusion technology could reduce image noise, improve signal-to-noise ratio, meanwhile improve definition and image quality, and be useful for detecting the embolus in pulmonary embolismObjective: To explore the value of dual energy technology and adjusted CTPA of dual-source CT in identifying emboli in pulmonary artery and pulmonary perfusion.Methods: 32 patients were examined by Germany Siemens dual- source CT in dual energy mode. The data were reconstructed by filtered back projection mode and 100 k V, 140 k V images and linear integration images whose mixing ratio was 3:2 were obtained. The data were reconstructed by 5 levels of sinogram-affirmed iterative reconstruction mode and 100 k V, 140 k V images and linear integration data whose mixing ratio was 3:2 were obtained. 100 k V,140 k V images of SAFIRE5 levels were reconstructed by Contrast Optimal software in dual-energy mode to generate a non-linear fusion images. Using 3D reconstruction software, linear integration images reconstructed by FBP mode and non-linear fusion images reconstructed by SAFIRE 5 levels mode were reconstructed.The pulmonary artery axial, sagittal and coronal images reconstructed by FBP mode were obtained, which was called traditional CTPA image. The pulmonary artery axial, sagittal and coronal images reconstructed by SAFIRE 5 levels mode were obtained, which was called adjusted CTPA image. Using pulmonary embolism detection technology(PED), non-linear fusion images reconstructed by SAFIRE 5 levels mode were reconstructed and pulmonary artery axial and curved image with pulmonary embolus were obtained(PED images). Using the lung vessels and lung PBV software, 100 k V,140 k V images reconstructed by SAFIRE 5 levels mode were reconstructed and color images of the blood, embolus in pulmonary arterial(Lung Vessels images)and pulmonary perfusion were obtained(DEPI images). Image’s quality of traditional CTPA images and adjusted CTPA images were scored by two doctors and the scores were analyzed by statistics software.Adjusted CTPA images, PED images, Lung Vessels images and DEPI images were diagnosed by two doctors in order. The location and number of the embolus in pulmonary artery and pulmonary perfusion abnormality were recorded. The consistency of diagnosis result of two doctors was evaluated by traditional CTPA images and adjusted CTPA images.The statistical difference of detection rates of traditional CTPA images and adjusted CTPA images were evaluated in observing the pulmonary artery branches, segment arteries, sub-segment arteries using the chi-square test. Then the consistency of diagnosis result of two doctors was evaluated by adjusted CTPA image, PED images, Lung Vessels images and DEPI images. The sensitivity and specificity of various kinds of reconstructions was calculated. The statistical difference of detection rates of CTPA image,PED image, Lung Vessels images and DEPI images were evaluated in observing the pulmonary artery branches, segment arteries, sub-segment arteries using the chi-square test. The consistency of changing of pulmonary perfusion displayed by DEPI+CTPA images and the embolus in pulmonary segment artery displayed by CTPA+PED+ lung Vessels images was evaluatedResults: The kappa of traditional CTPA images and adjusted CTPA images diagnosed by two doctors was 0.79.The scores of adjusted CTPA images were higher than the scores of traditional CTPA images(p <0.05). The detection rates of traditional CTPA images and adjusted CTPA images were not statistically different on diagnosis of pulmonary embolus in pulmonary artery branches, lobar artery, segment arteries and sub-segment arteries.(p >0.05)The kappa of CTPA image was 0.72, the kappa of CTPA+PED image was 0.94, the kappa of Lung Vessels+CTPA image was 0.9. The sensitivity and specificity of CTPA+PED images were the highest, respectively were 99.4% and 100%. The sensitivity and specificity of CTPA images were77.8% and 100%.The specificity of CTPA+lung vessels images and CTPA+PED images were 100%.But the specificity of lung vessels images and PED images 94.9%.93.4%. The detection rates of CTPA images, CTPA+PED images, CTPA+ Lung vessels images in diagnosing embolus of pulmonary arterial totally were statistically different each other.(p<0.05).The detection rates of CTPA images, CTPA+PED images, CTPA+Lung vessels images in diagnosing embolus of pulmonary artery branches and lobar artery were not statistically different each other.(p>0.05)The detection rates of CTPA images, CTPA+PED images in diagnosing embolus of segment arteries were statistically different.(p<0.05)The detection rates of CTPA images, CTPA+ Lung vessels images in diagnosing embolus of segment arteries were statistically different.(p<0.05)But the detection rates of CTPA+PED images, CTPA+ Lung vessels images in diagnosing embolus of pulmonary segment arteries were not statistically different.(p>0.05) The detection rates of CTPA images,CTPA+PED images, CTPA+ Lung vessels images in diagnosing embolus of pulmonary sub-segment arteries were statistically different each other.(p<0.05)The sensitivity and specificity of DEPI images in detecting abnormal of pulmonary perfusion were 92.1% and 96.5%. The sensitivity and specificity of DEPI +CTPA images were 92.1% and 100%.The kappa of DEPI images and DEPI +CTPA images compared with CTPA+PED+ lung vessels images were 0.85 and 0.95. In 32 cases of patients with PE, there was 70 pulmonary perfusion changing(sparse or defect), the perfusion sparse was 44cases; the perfusion defect was 26 cases. Pulmonary emboli were detected in 6cases, but there was no obvious perfusion abnormality. Pulmonary perfusion changes were detected in 14 cases; there was no obvious embolus in pulmonary artery.Conclusion: The technique of adjusted CTPA could improve the quality of images, but it dose not have impacts on the detection rate on general dose. The technique of pulmonary embolism detection and lung vessels could improve the efficiency of diagnosis and increase the sensitivity on diagnosis of pulmonary embolus in pulmonary segment arteries and sub-segment arteries. The technique of lung pulmonary blood volume could display the pulmonary perfusion. Joint application of adjusted CTPA could reduce the interference around and improve the diagnosis specificity. Joint application of dual energy technology and adjusted CTPA of dual-source CT were more valuable on Diagnosis of pulmonary embolism.