Objective To explore the application value of deep learning image reconstruction(DLIR)with coronary computed tomography angiography(CCTA)using 100 kV.Methods Sixty patients who underwent CCTA were selected.The tube voltage of 100 kV,noise index of 24 were applied.The 60％adaptive statistical iterative reconstruction-Veo(ASIR-V)and DLIR-low(DLIR-L),DLIR-medium(DLIR-M)and DLIR-high(DL1R-H)were reconstructed.The CT values and standard deviation(SD)values of the aortic root,left main artery,left anterior descending artery,left circumflex artery,right coronary artery and pericardial fat of the four groups of images were measured,and the signal-to-noise ratio(SNR)and contrast-to-noise ratio(CNR)were calculated.Two radiologists with five-year working experience subjectively evaluated the image quality using a five-point method double-blindly.Results The differences in noise(SD values),SNR values and CNR values among the four groups of images were statistically significant(P＜0.001).As the reconstruction gradually changed of 60％ASIR-V,DLIR-L,DLIR-M and DLIR-H,the coronary SD values gradually decreased,while the SNR values and CNR values gradually increased,among which the DLIR-H had the lowest SD values and the highest SNR values and CNR values.The subjective scores of the four groups of images by two radiologists showed good consistency(Kappa value=0.929,P＜0.001),and the subjective scores were all above 3 points which met the clinical diagnosis criteria.The subjective scores of DLIR-L,DLIR-M and DLIR-H were significantly higher than those of 60％ASIR-V(P＜0.001),with the DLIR-H achieving the highest subjective score.Conclusion DLIR can significantly reduce image SD value and improve image quality of CCTA with 100 kV,among which DLIR-H has the best effect on improving CCTA image quality.

Purpose To investigate the feasibility of CT-based feature tracking technology to quantify left ventricular myocardial strain(MS)and its significance in hypertrophic obstructive cardiomyopathy(HOCM).Materials and Methods A total of 35 HOCM patients who underwent cardiac coronary angiography from March 2019 to December 2021 in the First Affiliated Hospital of the Air Force Military Medical University were retrospectively included,and a total of 60 cases who were negative for cardiac coronary angiography among those who visited the hospital with suspected coronary artery disease were randomly enrolled.Conventional cardiac functional parameters and MS parameters were quantified via post-processing software,and differences of parameters between the groups were analyzed.The diagnostic efficacy of MS parameters for HOCM was further evaluated.Results Compared to the control group,the HOCM group exhibited significant increases in various conventional left ventricular functional parameters,including left ventricular wall thickness,mass,mass index,end-diastolic volume and stroke volume(t=2.119 to 24.861,all P<0.05).However,there were no statistically significant differences in end-systolic volume and cardiac output between the two groups(P>0.05).The global longitudinal and radial strain values of HOCM group were significantly lower than those of control group(t=12.857,-6.427,P<0.01),while the endocardial global circumferential strain of HOCM group was significantly higher than that of control group(t=-2.369,P<0.05).Among MS parameters,global longitudinal strain exhibited the best diagnostic efficacy for HOCM,with an area under the curve of 0.997.A cutoff value of≤20.78%for global longitudinal strain showed that the sensitivity and specificity was 100%and 95%,respectively.Conclusion The MS parameters quantified by the CT-based feature tracking technique are superior to left ventricular ejection fraction in quantifying left ventricular function,with the highest sensitivity and specificity for early myocardial function impairment of longitudinal strain.In addition,the technique has good repeatability and is expected to become a new indicator for the assessment of myocardial function in HOCM.

Objective To establish a neural network model based on enhanced CT for distinguishing ISUP grade of clear cell renal cell carcinoma (ccRCC). Methods We collected 131 cases of ccRCC, with 92 cases of low ISUP grade and 39 cases of high ISUP grade. Patients were divided into training set and validation set according to 5:5 stratified sampling. The enhanced CT images of each ccRCC patient were evaluated by the radiologist. Recursive feature elimination (RFE) was used to reduce the dimension of patients' general features and enhanced CT features, which was used for neural network modeling and validation. Results Patients' general features and enhanced CT features were verified by RFE method and then reduced to 14 features. The top 5 features were growth pattern, necrosis, enlargement of lymph nodes, tumor size and capsule. The AUC of the neural network model based on these 5 features in training set was 0.8844 (95%CI: 0.8062-0.9626), sensitivity was 89.47% and specificity was 82.61%; and those in validation set were 0.7924 (95%CI: 0.6567-0.9280), 75.00% and 86.96%, respectively. Conclusion The neural network model of ccRCC ISUP grade based on enhanced CT has relatively high diagnostic efficiency.

Objective: A predictive model of WHO/ISUP grading of renal clear cell carcinoma was constructed based on CT radiomics. Methods: The clinical data of 104 patients with ccRCC confirmed by operation or biopsy from March 2014 to December 2018 in the Affiliated Hospital of Shaanxi University of Traditional Chinese Medicine were retrospectively analyzed. There were 70 males and 34 females, and the age was 61.2±11.7 years. The patients were randomly divided into development cohort (73 cases) and validation cohort (31 cases) by stratified sampling according to 7∶3 ratio. According to the WHO/ISUP pathological grading criteria of renal cancer in 2016, Ⅰ and Ⅱ were defined as low-grade group, Ⅲ and Ⅳ were defined as high-grade group. The radiomics features of ccRCC were calculated in cortical phase images of CT enhanced scanning. LASSO regression was used to reduce the radiomics feature dimensionality in the training group, and to establish radiomics risk scores. The binary logistic regression was used to build the prediction model, which was used in the validation group. Bootstrap method was used to validate the model of training and validation group. AUC, sensitivity and specificity were calculated respectively. Hosmer-Lemeshow goodness-of-fit test was used to evaluate model calibration degree. Results: After dimensionality reduction, the radiomics risk score of ccRCC was established. The low and high-level risk scores of the training group were -2.49±1.73 and 1.23±2.17, with significant difference (t=-7.785, P < 0.01). The binary logistic regression multivariate analysis showed that the radiomics risk score was an independent risk factor in identifying low or high-grade ccRCC with odds ratio of (OR=3.576, 95%CI 1.964~6.513). The predictive model was Y=1/[1+ exp(-Z)], Z=1.274×radiomics risk score+ 0.072. The AUC of radiomics risk score in training group was 0.940 (95%CI 0.883-0.998) with 95.5% sensitivity and 88.2% specificity after internal verification by Bootstrap method, and good Hosmer-Lemeshow goodness-of-fit test (χ²=4.463, P>0.05). The low and high-level risk scores of the Validation group were -2.27±2.02 and 0.82±2.08, with significant difference (t=-3.832, P<0.01). The AUC in validation group was 0.859(95%CI 0.723-0.995) with 77.8% sensitivity and 81.8% specificity, and with good Hosmer-Lemeshow goodness-of-fit test (χ²=14.554, P=0.068) as well. Conclusions The prediction model based on CT radiomics has high accuracy in predicting high or low grade of ccRCC.

Objective To explore the clinical value of adaptive statistical iterative reconstruction(ASIR)combined with automatic tube current modulation in low dose scan of chest CT.Methods 80 patients who underwent chest CT scan were randomly divided into four groups,with a noise index(NI)of 14,16,18,and 20,respectively.Automatic tube current modulation technique for chest CT scan was adopted for all patients.The standard dose group,with a NI of 14,was reconstructed with FBP.While the low-dose groups,with a NI of 16,18 and 20,were reconstructed with four ASIR levels(20% ASIR,40% ASIR,60% ASIR and 80% ASIR) in each group.The volume CT dose index(CTDIvol),dose length product(DLP)and effective dose(ED)were recorded of the four groups.The signal-to-noise ratio(SNR)and contrast-to-noise ratio(CNR)were calculated of the different reconstruction groups. Two radiologists with more than five years of work experience blindly scored the subjective image quality.Results Effective dose of the four groups with different NI were(3.29 ± 1.21)mSv,(3.10 ± 1.07)mSv,(2.20 ± 0.82)mSv and(1.97 ± 1.21)mSv,respectively. In all the reconstruction groups,when NI was constant,the greater the ASIR,the lower the SD.When ASIR percentage was constant, the SD was increased and the SNR was reduced along with NI rising up.When the parameters were set as NI 18 and ASIR 60%,the SD would be smaller and SNR would be higher than that of a standard dose group.In aspect of the scores from subjective image quality evaluation,there was no significant difference(P> 0.05).The average ED was decreased by about 33.2% when compared to the standard dose group.Conclusion Combining automatic tube current modulation with ASIR has a certain clinical practicality.When the NI is set at 18 and ASIR is 60%,the image quality can be optimal for not only satisfying the requirements of clinical diagnosis, but also reducing radiation dose in chest CT scanning.

Objective To demonstrate the feasibility of high-resolution computed tomography(HRCT)reconstructed with a model-based iterative reconstruction (MBIR)for evaluating early peripheral lung cancer (≤3 cm),by comparing image quality obtained from MBIR,filtered back proj ection reconstruction(FBP)and state of the art adaptive statistical iterative reconstruction(ASIR)algorithm respectively.Methods A total of 30 patients confirmed with lung cancer by postoperative pathology were enrolled in the study.A chest phantom was also used to evaluate image noise,spatial resolution and density resolution.Both patients and chest phantom were received HRCT,and the images were reconstructed using FBP,ASIR(40% ASIR and FBP mix)and MBIR.The objective CT value, standard deviation(SD)and signal noise ratio(SNR)were measured.Two radiologists used a semi-quantitative to rate subjective image quality of lung nodules.Results There was no significant difference in CT value between the three reconstruction algorithms (P>0.05).But significant improvements in objective image noise were observed in MBIR compared with FBP and ASIR (P<0.05):including the SD value in back muscle [(12.63±1.70)with MBIR vs (31.58±5.21)with FBP and (24.55±4.14)with ASIR],and in subcutaneous fat [(12.77±2.53)vs (24.39±5.08)and (19.20±4.11)].Subjective image noise of the three group were also significantly difference:FBP with lowest subjective noise score;and MBIR with highest subjective noise score.The sharpness of small vessels and bronchi and diagnostic acceptability with MBIR were significantly better than with FBP and ASIR (P< 0.05).Conclusion Lung HRCT reconstructed with MBIR provides diagnostically more acceptable images for the detailed analyses of peripheral lung cancer compared with FBP and ASIR.

Objective To compare the spatial resolution and density resolution balance algorithm(MBIRSTND)and spatial resolution preference algorithm (MBIRRP20)from new version of model-based iterative reconstruction(MBIRn),and adaptive statistical iterative reconstruction(ASIR) with lung kernel in routine dose about the performance of computer-aided detection (CAD)for quantitative analysis of airway.Methods 30 patients were involved who were scanned for pulmonary disease with spectrum CT.Data with a slice thinkness 0.625 mm were reconstructed with ASIR,MBIRSTNDand MBIRRP20.Airway dimensions from three reconstruction algorithm images were measured using an automated and quantitative software(Dexin-FACT)that was designed to segment and quantify the bronchial tree,and a skeletonization algorithm to extract the center-line of airway trees automatically.For each patient,reconstruction algorithm chose the right middle lobe bronchus,and the bronchial length of the matched airways was measured by this scheme.Two radiologists used a semiquantitative 5 scale (Score 0 stands for its image quality is similar to that with ASIR;Score±1 stand for a little better or a little worse;Score±2 stand for obviously better or obviously worse)to rate subjective image quality of airway trees about images reconstructed with MBIRSTNDand MBIRRP20.Paired t test and Wilcoxon signed-rank test were used.Results Algorithm impacts the measurement variability of bronchus length in chest CT.The bronchial length with MBIRRP20was longer than with MBIRSTND, while the length with ASIR were the shortest(P<0.05).In addition, the optimal reconstruction algorithm was found to affect the subjective noise,the continuity and completeness of bronchial wall,and the show of bronchial end.The subjective noise of MBIRSTNDwas better than that of MBIRRP20.The show of bronchial end of MBIRRP20was better than that of MBIRSTND(P<0.05).There was no significant difference in the continuity and completeness of bronchial wall compared with MBIRRP20and MBIRSTND(P>0.05),which was much better than with ASIR(P<0.05).Conclusion MBIRn can inmprove the analyzing ability of CAD airway.The MBIRSTNDcan significantly reduce the image noise,the MBIRRP20significantly improve the branching of the bronchial arteries,both of which can allow the desired airway quantification accuracy of CAD for chest CT of the bronchial wall.

Objective To evaluate the image quality improvement on abdominal CT imaging by using new model-based iterative reconstruction (MBIRn)in comparison with adaptive statistical iterative reconstruction (ASiR).Methods 40 patients who underwent upper abdominal three-phase contrast-enhanced scan were included.After scanning,all the scans obtained at 180 s later injection were reconstructed by three protocol,including ASiR (combined reconstruction of 40%FBP and 60%ASiR),the MBIRn of the noise reduction settings (MBIRNR40)and the spatinal resolution settings(MBIRSTND).The thickness of the slice was 0.625 mm.The values of CT and SD of the subcutaneous fat,left erector spinae,inferior vena cava and hepatic vein (left and right branches)were measured at the branch level of hepatic vein,and the contrast noise ratio (CNR)between inferior vena cava and hepatic vein were calculated.The subj ective image quality was evaluated by two radiologists according to the noise,smoothness and small branches of the inferior vena cava and hepatic vein using 5-scoring method.The quality images obtained from ASiR method were treated as reference standard.Results For MBIRSTNDand MBIRNR40images,the subjective noise decreased and image quality increased comparing with ASiR images.Among which the MBIRNR40images had the best image with vascular smoothness score and the lowest subjective noise.Conclusion Compared with ASiR,MBIRSTNDand MBIRNR40,especially MBIRNR40improves the quality of CT images of the inferior vena cava and its branches.

Objective To explore the value of intravoxel incoherent motion DWI (IVIM-DWI) combined with single-voxel MRS in distinguishing osteoporotic fractures from metastatic vertebral compression fractures.Methods Totally 70 patients with vertebral compression fractures,who underwent CT scanner were enrolled.The patients were divided into osteoporotic group or metastatic group based on pathological results or clinical follow-up.All patients underwent conventional sagittal T1W,T2W,STIR,IVIM-DWI and single-voxel MRS scanning.Relative peak areas of the signal of water at 4.7 ppm and lipid at 1.3 ppm were determined.IVIM-DWI parameters (diffusion coefficient [D],pseudo diffusion [D*],perfusion fraction [f]) and MRS parameters (lipid water ratio [LWR],fat fraction [FF]) were also recorded.The diagnostic performance of MRS,IVIM-DWI,as well as MRS combined with IVIM-DWI in distinguishing osteoporotic fractures from metastatic vertebral compression fractures were evaluated by using ROC curve,and the area under curve (AUC) was calculated.Results The f,D and FF in metastatic group were significantly lower than those in osteoporotic group,while D* in metastatic group was significantly higher than that in osteoporotic group (all P＜0.05).The sensitivity,specificity and accuracy in differentiating osteoporotic fractures from metastatic vertebral compression fractures was 87.50％ (28/32),57.89 (22/38) and 71.43％ (50/70) of MRS,78.13％ (25/32),89.47％ (34/38) and 84.28％ (59/70) of IVIM-DWI,90.63％ (29/32),97.37％ (37/38) and 94.29％ (66/70) of MRS combining with IVIM-DWI,respectively.AUC of MRS,IVIM-DWI,as well as MRS combined with IVIM-DWI was 0.73,0.88 and 0.94 (all P ＜0.05),respectively.Conclusion Combination of IVIM-DWI and MRS can improve the diagnostic efficiency of differentiating osteoporotic fractures from metastatic vertebral compression fractures.

Purpose To explore the value of spectral coronary CT angiography (CCTA) in reducing radiation dose and contrast dose using individualized optimal monochromatic imaging.Materials and Methods Sixty patients with suspected coronary disease were recruited in the study,who were randomly divided into two groups:group A (n=30) using conventional CT protocol with 350 mgI/ml contrast agent;group B (n=30) using low dose spectral CT imaging mode with 300 mgI/ml contrast agent.The images of group A were reconstructed with conventional process,and the images of group B were reconstructed with Optimal CNR to obtain the optimal monochromatic energy images.The images of both groups were transferred to an Advanced Workstation for analysis.Double-blinded method was carried out to qualify the images.CT values of coronary artery segments,as well as standard deviations (SD),the signal-to-noise ratio (SNR),contrast-to-noise ratio (CNR) of aortic sinus were measured.Radiation doses and iodine intake were compared between the two groups.The optimal keV distribution in group B was analyzed.Results There was no significant difference for the subjective scoring of image quality,CT value of each coronary artery segment,SD,SNR,and CNR values between the two groups (P＞0.05).The effective radiation dose and total iodine load in group B were less than that in group A (P＜0.05).The optimal energy distribution for group B was 60-75 keV,average at (66.50+3.91) keV.Conclusion Compared with the conventional CT protocol,spectral CT imaging at optimal energy levels combined with iterative reconstruction can effectively reduce the radiation dose and iodine load,and obtain better images than usual protocol.