Font Size: a A A

Spectral CT Imaging Of The Portal Vein

Posted on:2016-10-19Degree:MasterType:Thesis
Country:ChinaCandidate:J ZhaoFull Text:PDF
GTID:2284330461476814Subject:Imaging and nuclear medicine
Abstract/Summary:PDF Full Text Request
Part ⅠComparison of portal vein image quality with spectrum CT based on different levels of Adaptive statistical iterative reconstruction algorithm (ASiR)[Purpose] To assess the quality of portal vein images using of spectral CT imaging and adaptive statistical iterative reconstruction (ASiR) technology with different levels of adaptive statistical iterative reconstruction algorithm, and investigate the optimization of ASiR algorithm in the proportion of portal vein imaging.[Materials and Methods] 40 patients had the history of tumor performed contrast-enhanced abdominal scanning using spectrum scanning mode and fast switching dual voltage. A recombinant single energy image 60keV energy section were selected from the scan data, respectively FBP (0%ASiR),30% ASiR,40% ASiR,50% ASiR, 60%ASiR algorithm. Images were reconstructed with five groups of different levels of ASiR algorithm. Single factor analysis of variance was used between the five groups for the image noise of portal vein and liver parenchyma. Two radiologists with more than 10 years working experience of abdominal diagnosis independently evaluated the five image quality score for the image noise, artifacts and the capacity of showing the tiny structures of three level portal vein branches.[Results] Compared FBP (0%ASiR) algorithms with 30% ASiR,40% ASiR,50% ASiR and 60%ASiR, the objective image noise in liver parenchyma and portal vein were gradually decreased, and the differences were statistically significant. In the images of 30% ASiR,40% ASiR,50% ASiR,60%ASiR algorithms, the objective image noise rate of decrease were 17.58%,23.97%,28.64%,34.47% in liver parenchyma and 16.58%,22.64%,28.15%,33.30% in portal vein. Compared 40%ASiR algorithms with 50% ASiR and 60% ASiR, the differences of the objective image noise in liver parenchyma and portal vein were not statistically significant. With the increase of ASiR value, the subjective score of five different kinds of reconstruction which got by two radiologists with more than 10 years working experience in image noise and artifacts were gradually increased, but the scores of showing the tiny structures of 3 degree portal vein branches in 40% ASiR algorithms were the highest, and the scores of 60% ASiR algorithms were the lowest. Between five different reconstruction algorithms, subjective score differences of the image noise, artifacts, capacity of showing the tiny structures of 3 degree portal vein branches were statistically significant (each P=0.000). In the display of 3 degree portal vein branches, compared 40% ASiR algorithms with 50% ASiR, 60%ASiR, the differences were statistically significant (P=0.000、0.000).[Conclusion] ASiR reconstruction algorithm can improve image quality by reducing image noise, with the increase of ASiR reconstruction ratio, image noise, artifacts are reduced, the image density resolution increased. But the high ASiR reconstruction algorithm itself will cause the image of "fuzzy" effects. When the ASiR value set to 40%, the image quality is the best, which can reduce the image noise, and not casuse obvious "fuzzy" effects on images.Part ⅡEvaluation of the portal vein images combiningmonochromatic spectral CT with adaptive statistical iterative reconstruction (ASiR)[Purpose] To investigate and evaluate the quantitative and subjective qualityof the portal vein images combining monochromatic spectral CT with adaptive statistical iterative reconstruction (ASiR).[Materials and Methods] 90 patients withtumorhistory routinely receiving abdomen CT scans were randomly divided into three groups.Group A (optimal monochromatic group) included 30 patients who underwent fast kVp-switching dual-energyabdominalenhanced CT,from which101 sets of portal phase recombinantmonochromatic(40~140 keV)images were generated, and the optimal monochromaticset with the highest contrast to noise ratio (CNR) were obtained. Group B (60keV+40%ASiR group) included 30 patients who underwent fast kVp-switching dual-energy abdominal enhanced CT, and portal phase 60keV recombinant monochromatic images were obtained and reconstructed with 40% ASiR. Group C (conventional multi-detector CT group) included 30 patients who underwent traditional abdominal enhancedmulti-detectorCT, whoseportal phase images were obtained.The average CT values of the portal vein and the hepatic parenchyma, as well as the portal-hepatic HU difference, the image noise and the CNR, were obtained and compared among the three groups using the single factor analysis of variance.2 board-certified radiologists interpreted the selected image sets independently for image quality scores.[Results] Themonochromatic images were found to yield the optimal CNR for portal vein at 55.87keV.In group A, B and C,the CNR were 6.81±1.87,8.09±2.29 and 3.30±1.08, and the image quality scores were 4.28±0.58,4.35±0.56 and 3.58±0.46, respectively.The CNR and image quality scores of group A and Bwere significantly higher than group C (all P<0.001).[Conclusion] Compared to conventional CT imaging, the 60keVmonochromatic spectral CT imaging with 40%ASiR and the optimal monochromatic CT imaging bothcan significantly improve the image quality of portal vein.
Keywords/Search Tags:Tomography, X-ray computed, Adaptive statistical iterative reconstruction, Filtered back projection, Spectral imaging, Portal vein
PDF Full Text Request
Related items