Objective To build a deep learning method based on CT for hematoma segmentation and hematoma expansion (HE) prediction of patients with hypertensive intracerebral hemorrhage (ICH). Methods This study retrospectively analyzed 500 cases of hypertensive intraparenchymal hematoma.Baseline non-contrast-enhanced CT scans (NECTs) were collected at admission and compared with subsequent CTs to determine the presence of HE.A deep learning method based on CT was proposed to automatically segment the hematoma region.In order to build the model for hematoma segmentation and HE prediction,patients were randomly divided into the training set and the test set according to the ratio of 7:3.The training set contains 350 patients for training the model,while the test set contains 150 patients for model validation.The segmentation module adopts the attention U-net to automatically segment and detect the hematoma,while the prediction module uses the resnet-34 to predict the HE.Mean intersection over Union (Miou) and Dice coefficient were used to evaluate the segmentation task.The accuracy evaluation of the prediction model was evaluated using the area under curve (AUC) of the receiver operating characteristic (ROC) curve.The continuous variables were compared by independent sample t-test or Mann Whitney U test.P<0.05 was considered statistically significant. Results There was no significant difference in age (P=0.211) and gender (P=0.213) between the HE and the non-HE groups.Hematoma volume (16.7 ml vs.26.7 ml,P< 0.001) and hematoma diameter (41.7mm vs.48.9mm,P< 0.001) in the non-HE group were significantly lower.The Miou and Dice coefficient of the segmentation model was 0.89 and 0.90 respectively.The AUC value of the prediction model was 0.909 (95%CI:0.85 -0.949),and the sensitivity and specificity were 81.58% (95%CI:65.7%-92.3%) and 96.43% (95%CI:91.1%-99.0%) respectively. Conclusion The deep learning system based on brain CT can accurately segment hematoma and predict the expansion of hematoma.

Objective To investigate the radiation performance of new domestic DSA equipment by comparing the radiation dose of two digital subtraction angiography devices during percutaneous vertebraplasty. Methods Patients undergo-ing elective percutaneous vertebraplasty were randomly assigned to the trial group [United Imaging uAngio(UIuA)] and the control group [Siemens Axiom Zee Biplane(SAZB)] after being matched by a 5% range of difference in their body mass index (BMI).Comparison and analysis of image quality evaluation,Fluoroscopy Time (T),Dose-Areas Product rate(DAP rate),Air Kerma rate (K rate),and Entrance Surface Dose (ESD) magnitude on the left and right sides of the patient's torso at the surgical site for the two devices. Results There was no significant difference between the fluoroscopic and spot image evaluations of the UIuA and the SAZB; The DAP rate of the UIuA (31.88; 24.01-44.15)μGy·mm²/min was significantly lower than that of the SAZB (63.73; 51.62-80.26μGy·mm²/min),P= 0.022.K rate was not significantly different between the UIuA (4.74; 3.17-5.97 mGy/min) and the SAZB (5.76; 4.58-8.73 mGy/min),P=0.074.The ESD was significantly lower in the right and left sides of patients with the UIuA than with the SAZB,both the P-values < 0.05. Conclusion On the premise of equal image quality evaluations,the domestic DSA device UIuA outperforms the SAZB in terms of radiation dose. UIuA is important for reducing radiation exposure and thus radioactive damage,as well as for improving the radiation safety environment for interventional medicine in China.

Objective To investigate the changes in topological properties of resting-state functional brain networks in alcohol-dependent patients and their relationship with cognitive function and clinical performance. Methods Resting-state functional magnetic resonance imaging scans were performed on 53 alcohol-dependent patients (AD group) and 68 age and education-matched healthy controls (HC group).The functional brain network model was constructed using graph theory to compare the small-world network characteristics and nodal properties of brain regions in the two groups,and to analyses the correlation between brain network changes and the Penn Alcohol Craving Scale (PACS),the Beck Anxiety Inventory (BAI),the Mini-Mental State Examination (MMSE),and the Digital Span Test (DST). Results The brain networks of both groups had small-world network characteristics (γ＞1,λ ≈ 1,σ＞1).The differences in γ and σ between the two groups were not statistically significant (P＞0.05).In the sparsity range of 0.20 to 0.40,λ was reduced in the alcohol-dependent group compared with the healthy control group,and the differences were statistically significant (P＜0.05).The alcohol-dependent group had higher degree centrality in the left insula and left amygdala,higher betweenness centrality in the left temporal transverse gyrus,and lower betweenness centrality in the left putamen compared with the healthy control group (all P＜0.05).The left insula was positively correlated with DST scores (r=0.273,P=0.048),and the left amygdala was positively correlated with PACS (r=0.322,P=0.019) and BAI (r=0.293,P=0.033) scores. Conclusion The information transmission capacity of functional brain networks in alcohol-dependent patients is impaired,and graph theory analysis of brain network can be used as an imaging indicator of the abnormal brain function in patients with alcohol dependence.

Objective To analyze the correlation between quantitative parameters of abdominal adipose tissue on CT and composition of carotid atherosclerotic plaque in patients divided by age. Methods The clinical and imaging data of 160 patients were retrospectively evaluated. All patients underwent abdominal CT and head-neck computed tomography angiography (CTA) examination,and the time interval of two examinations was less than one month. All raw data of original abdominal CT were uploaded to postprocessing workstation (GE AW4.7).Visceral fat area (VFA) and subcutaneous fat area (SFA) were measured at umbilical level on workstation by one radiologist. Then,the ratio of visceral to subcutaneous fat area (VFA/SFA) was calculated according to the results of VFA and SFA. Carotid plaques were evaluated according CTA by one senior radiologist. According to the American Heart Association (AHA) guidelines,soft plaques and mixed plaques mainly composed of fat were classified as vulnerable carotid plaques while calcified plaques were classified as stable plaques. Patients were divided into age < 70 years subgroup and age ≥ 70 years subgroup. The differences of clinical characteristics and quantitative parameters of abdominal adipose tissue were separately compared between stable plaque group and vulnerable plaque group. Multivariate Logistic regression analysis was used to analyze the risk factors of vulnerable carotid plaque. ROC curve analysis was used for indicating the predictive performance of different variables for vulnerable carotid plaque. P value less than 0.05 was considered statistically significant difference. Results In the subgroup of patients aged < 70 years subgroup,the proportion of patients with VFA > 151.65 cm2 in the vulnerable plaque group was higher than that in the stable plaque group (P= 0.019).Nevertheless,there was no significant difference in VFA and SFA between the two groups (P= 0.523,0.288).In the subgroup of patients aged ≥ 70 years subgroup,gender and SFA showed significant difference between vulnerable plaque group and stable plaque group (P= 0.005,0.003).Then,gender,SFA and VFA/SFA were included in multivariate analysis (P< 0.1).Multivariate analysis showed that compared with patients with SFA ≤ 156.27 cm²,the OR value of patients with SFA > 156.27 cm² was 0.264 (95% CI = 0.079-0.887,P= 0.031),with statistical significance. Male (OR = 3.676,95% CI:1.164-11.609,P= 0.026) was an independent risk factor for patients with vulnerable carotid plaque in aged ≥ 70 years subgroup. The AUC under SFA > 156.27 cm² for predicting vulnerable carotid plaque were 0.684,respectively. Conclusion VFA was associated with composition of carotid atherosclerotic plaque in patients aged < 70 years. Male or patients with lower SFA had a higher risk of vulnerable carotid plaque in patients aged ≥ 70 years.

Objective To develop a CESM model for predicting malignant probability of breast masses,and to evaluate its predictive. Methods The clinical and imaging data of 328 patients were analyzed retrospectively,which were randomly divided into training set (235 cases) and verification set (93 cases).To pathology results for the gold standard,the predictors were screened by univariate and multivariate Logistic regression analysis. Then a predictive model was constructed based on the results with a nomogram drawn. Performances of predictive models were evaluated with area under the curve (AUC) of ROC,calibration curve,and decision curve analysis (DCA). Results A total of 328 women with breast neoplasia including 179 malignant and 149 benign was collected. Six predictive factors were harvested to construct the predictive model,which included age,lobulation,spiculation,early-phase (CC) enhancement degree,late-phase (MLO) enhancement heterogeneity or enhancement curve. The predictive model had good performance in both training set and verification set,with the AUC values of 0.975 and 0.963,respectively. The calibration and DCA curve showed that the predictive result was much closer to actual result and this model had good clinical application values. Conclusion The CESM predictive model can predict malignancy probability of breast masses,which can be used as a referable method for radiologist and to provide guidance for clinical practice.

Objective To evaluate the value of multi-slice spiral CT in the differential diagnosis of calcification of the ligamentum arteriosum (CLA) and aortic arch calcified plaque in the presenile population. Methods 200 patients aged 50-59 years with clinically suspected chest lesions were retrospectively analyzed,scanned with multi-slice spiral CT,calculated the incidence,observed the location and the shape,measured the density of CLA and aortic calcified plaque in the thin-layer mediastinal window image. Results CLA occurred between the aortic arch and the left pulmonary trunk,most of which were strip and spotted,with an incidence of 27.00%,the CT value was (247.63±92.23) HU. Aortic calcified plaque occurred in all positions of the aortic arch,patchy and spotted,some were multiple,with an incidence of 45.00%,and the CT value was (448.44±222.41)HU. Conclusion CLA occurs in the area of arterial ligament,mostly in strip and spotted shape,which is significantly different from aortic calcified plaque in incidence,location,shape and density,et al.

Objective To investigate the value of Gd-EOB-DTPA MRI T1 mapping in the evaluation of liver function. Methods This study prospectively collected the clinical and imaging data of 42 patients who met the inclusion and exclusion criteria,and divided them into three groups according to Child-Pugh classification:23 cases in the Child-Pugh grade A (LCA group),13 cases in the Child-Pugh grade B (LCB group),and Child-Pugh grade C ( LCC group) 6 cases. The T1 relaxation time (T1_pre) and signal intensity (SI) of the liver before enhancement were measured,the liver T1 relaxation time (T1_HBP) and SI of hepatobiliary phase were measured; the SI ratio of liver/erector spinae (SIpre,SIHBP) before enhancement and hepatobiliary phase were calculated,respectively; the reduction rate of liver T1 relaxation time (△T1%) and relative liver enhancement rate (RLE) were calculated. The difference between SI and T1 quantitative parameters between different groups were compared,and the performance of different parameters in evaluating Child-Pugh B+C grade were analyzed. Pearman correlation coefficient was used to analyze the correlation between each parameter and indocyanine green retention rate at 15 min (ICG R15). Results The area under the curve of ROC of T1_pre,T1_HBP,Δ T1%,SIpre,SI_HBP,and RLE diagnosing Child-Pugh B+C grade were 0.714,0.884,0.914,0.643,0.832,0.834,respectively.T1_HBP was significantly positively correlated with ICG R15 and Child-Pugh scores (r= 0.677,r= 0.662,P<0.05). Conclusion T1_HBP is not only significantly positively correlated with ICG R15 and Child-Pugh scores,but the diagnostic performance is better than ΔT1% and the correlation parameter of SI,which has an important value for liver function assessment.

Objective To investigate the value of tumor blood flow imaging (TBF) with three-dimensional arterial spin labeling (3D-ASL) perfusion imaging for the assessment of non-enhancing cranial astrocytoma grading.Methods Twenty-five patients with non-enhancing cranial astrocytomas who presented to our hospital between March 2020 and March 2022 and were pathologically confirmed were retrospectively collected as study subjects and divided into high-grade group (n=15) and low-grade group (n=10) according to tumor grading.The absolute maximum TBF (TBF_max),absolute mean TBF (TBF_mean) and the corresponding normalized values of cerebral blood flow (TBF_max rate and TBF_mean rate) of the tumor were calculated after the enrolled subjects were scanned by MRI conventional sequences and ASL.The differences in TBF_max,TBF_mean,TBF_max rate,and TBF_mean rate between the two groups of subjects were compared using independent samples t-test,and the subject operating characteristic curve (ROC) was applied to predict the accuracy of each index for tumor grading assessment.Results ROC curve analysis showed that TBF_max rate had the highest accuracy in predicting non-enhancing high-grade cranial astrocytoma (AUC=0.960,best cut-off value of 1.35,Youden of 0.800,sensitivity 80.00%,and specificity 100.00%),TBF_mean rate (AUC=0.887,best cut-off value of 1.36,Youden index of 0.600,sensitivity 80.00%,specificity 80.00%) and TBF_max (AUC=0.853,best cut-off value 76.01 ml/100 g/min,Youden index 0.633,sensitivity 73.33%,specificity 90.00%) were next,and TBF_mean (AUC= 0.773,best cutoff value of 30.09 ml/100 g/min,Youden of 0.533,sensitivity 93.33%,specificity 60.00%) was the lowest.Conclusion The 3D-ASL perfusion imaging parameters,especially the TBF_max rate,can more accurately distinguish between high-grade and low-grade non-enhancing cranial astrocytomas,which can provide a sufficient theoretical basis for individualized clinical treatment plans.

Objective The aim of this study was to assess sex- and age-specific normal values of left atrium (LA) myocardial strain (MS) and strain rate (SR) in healthy adults in China by cardiac magnetic resonance (CMR). Methods 110 healthy adults (45.65±12.77,76 males) were retrospectively included,all of whom underwent 3.0T CMR and were stratified according to gender and age.LA global longitudinal strain (GLS),GLS SR,and global circumferential strain (GCS) were assessed using CMR.The GLS is divided into three stages:Reservoir,Conduit and Booster. Results LA GLS Reservoir,GLS Conduit,GCS and SR Reservoir were significantly greater in females compared to males(P<0.05).Differences in LA GLS Reservoir,SR Reservoir and GLS Conduit and SR Conduit between age groups were statistically significant (all P<0.05) and gradually decreased with increasing age.Among females,age was significantly negatively correlated with LA GLS Conduit,SR Conduit (P<0.001),significantly positively correlated with GLS Booster (P=0.034) and not significantly correlated with GLS Reservoir (P=0.070).In the male population,a significant negative correlation (P≤0.001) could be found between age and both LA GLS (Reservoir,Conduit) and SR (Reservoir,Conduit). Conclusion CMR is feasible and reproducible for quantifying LA MS and SR and determining left atrial phase function.The presence of age- and sex-related differences in LA MS and SR suggests the need to establish specific normal values for individual groups.

Objective To explore the clinical value of dynamic volume perfusion CT (dVPCT) for multidimensional evaluation of T-staging and pathological grading of pancreatic carcinoma (PC). Methods 56 cases of pathologically confirmed PC who underwent preoperative epigastric dVPCT in Deyang People's Hospital were retrospectively collected from July 2017 to March 2022.Based on multiple reconstructive morphological images,the PC was observed and preoperatively T-staged.The Cohen’s weighted Kappa (Kw) concordance analysis was used to compare dVPCT T-staging with postoperative pathological T-staging and the accuracy of dVPCT for preoperative T-staging of PC was assessed.Blood flow (BF),blood volume (BV),permeability surface(PS) and mean transit time (MTT) in each case were measured.Patients were divided into T1-2,T3,and T4 groups,and 31 of them were classified into highly,moderately,and poorly differentiated groups based on pathologic stage.The Kruskal-Wallis H test was used to compare the perfusion parameters of PC with different T stage and different pathologic stage. Results The detection rate of PC lesion was 100%.dVPCT staging was remarkably consistent with pathological T staging (Kw 0.83 ± 0.07,95% CI 0.70-0.96).The T staging accuracy was 87.5%,with T1,T2,T3 and T4 staging accuracy were 96.4%,89.3%,92.9% and 96.4%,respectively.BF,BV,PS and MTT were significantly lower in T4 than in T1-2 (P＜0.05).BV,PS and MTT were significantly lower in T3 than in T1~2 (P＜0.05).The BF,BV and PS values in the poorly differentiated group were significantly lower than those in the highly differentiated group (P＜0.05).The BF and BV values in the poorly differentiated group were significantly lower than those in the moderately differentiated group (P＜0.05). Conclusion dVPCT can acquire both high quality images and functional parameters,which is more clinically useful for PC display,T-staging and pathological grading.

Objective To observe the value of radiomics based on different regions of interest on multiparametric MRI for the assessment of prostate cancer aggressiveness. Methods In this retrospective study,patients with pathologically confirmed prostate cancer from two different centers were included,with center 1 including 123 cases as a training cohort and 53 cases as an internal validation cohort.Center 2 included 69 patients as an external validation cohort.Two different segmentation methods were used on multiparametric MRI:segmentation of prostate region and lesion,and after feature extraction and reduction,radiological feature sets were obtained.Using logistic regression algorithms,three radiomic models were established based on three radiomic feature sets:model 1 (based on prostate region),model 2 (based on lesion region),and model 3 (based on prostate + lesion).Clinical features associated with prostate cancer aggressiveness were analyzed to create an integrated model combining radiomic features with clinical features (model 4).Receiver operating characteristic (ROC) curve analysis was used to compare the performance of each model in assessing prostate cancer aggressiveness. Results The region of interest based on the prostate region retained the 57 best features,the region of interest based on the lesion region retained the 27 best features and the region of interest based on the prostate region + lesion retained the 64 best features.The model with both prostate and lesion region segmented (AUC 0.92) performed better in assessing prostate cancer aggressiveness compared to the model with only prostate region segmented (AUC 0.89) and the model with only lesion region segmented (AUC 0.87),and was comparable to the performance of the integrated model combining radiomic and clinical features (AUC 0.92). Conclusion A radiomics model combining the prostate region and the lesion region may perform better in assessing prostate cancer aggressiveness than a radiomics model segmenting only the lesion region or segmenting only the prostate region.

Objective This study aims to investigate hemodynamic alterations in saccular intracranial aneurysms and their adjacent arteries using four-dimensional flow magnetic resonance imaging (4D-FLOW MRI). Methods We enrolled 68 patients diagnosed with intracranial aneurysms through CT angiography and head MR angiography at Wuhan Union Hospital. Hemodynamic analyses were conducted after 4D-FLOW MRI scanning and reconstruction,and the hemodynamic parameters were automatically derived through the manual delineation of vessel contours at various critical locations within the parent artery. These locations included the aneurysm inlet cross-section,the aneurysm's largest cross-section in alignment with the blood flow direction,the aneurysm outlet cross-section,and symmetrical positions corresponding to the contralateral arteries. Paired-sample t-test was employed to compare hemodynamic parameters between the parent artery and the contralateral artery. According to the grouping based on the maximum diameter and size ratio (SR) of saccular aneurysms,an independent sample t-test was employed to analyze the differences in hemodynamic parameters of the parent arteries among different groups. Results In the parent artery compared with the contralateral artery,the maximum WSS,mean WSS,and maximum axial WSS-maximum circumferential WSS at the entrance of the intracranial aneurysm were smaller than those in the contralateral symmetrical cross section; the maximum WSS,mean WSS,mean axial WSS,and mean axial WSS-maximum circumferential WSS at the largest cross section of the aneurysm along the direction of blood flow were smaller in the parent artery,and maximum energy loss and mean energy loss were significantly higher in the contralateral arteries,and all these differences were statistically significant(P<0.05).In the parent arteries,the maximum axial WSS and mean axial WSS of the aneurysm inlet section was bigger in the group with a maximum diameter of >5 mm than in the group with a maximum diameter of ≤5 mm,and the mean axial WSS of the aneurysm inlet section was greater in the SR>2 group compared with that of the SR≤2 group,and the differences between the groups were statistically significant(P<0.05). Conclusion The application of 4D-FLOW MRI can dynamically display the flow environment in the intracranial saccular aneurysm and the parent artery around the aneurysm neck,and quantitatively evaluate the hemodynamic characteristics,which is of great value in exploring the formation factors of intracranial saccular aneurysms.

Objective This study aimed to explore the correlation between ring wall enhancement types of Glioblastoma and inflammation by quantitative and semi-quantitative analysis of dynamic contrast enhancement. Methods Twenty-three GBM patients and seventeen Inflammatory lesion patients (eleven brain abscess and six tuberculomas),who confirmed by pathology,were retrospectively analyzed between September 2015 and September 2020.The tumor wall of GBM was composed of two types:thick and irregular ground glass enhancement and thin and smooth thin wall enhancement. For inclusion in this study,patients need to have undergone a chest 3.0 T MRI examination within 2 weeks before surgery. MRI examination include T₁-weighted MRI,T₂-weighted MRI,T₁ enhancement MRI and DCE. The region of interest (ROI) of the largest layer of tumor and inflammatory lesions was manually delineated on DCE images using GE Omni-Kinetics Version software. The DCE parameters of thick wall ground-glass enhancement pattern(group 1),thin wall enhancement pattern(group 2) and inflammatory lesion ring enhancement pattern(group 3) were compared. Including the difference of volume transfer constant (K^trans),volume of the extracellular space (V_e),area under curve (AUC),back flux Constant (K_ep),plasma volume fraction (V_p),time to peak (TTP) and maximum rise slope (MAXslope).K^trans,AUC and V_p were compared between the three groups with use of the variance analysis and least significant difference (LSD).V_e,K_ep,TTP and MAXslope were compared between the three groups with use of the rank sum test and Kruskal-Wallis test. Results There were significant differences in K^trans,AUC and MAXslope among the three groups (P< 0.05).The K^trans,AUC,and MAXslope of the thick-walled ground-glass enhancement area of GBM were higher than those of the thin-walled enhancement area and the ring-wall enhancement of inflammatory lesions. There were no significant differences in K^trans,AUC and MAXslope of the latter two areas (P> 0.05). Conclusion The thin-walled enhancement area of GBM has similar hemodynamic and permeability characteristics to the circumferential wall enhancement area of inflammatory lesions,and it is speculated that the inflammatory process may be involved in the thin-walled enhancement area of GBM.

Objective To analyze the MR and CT imaging manifestations of central nervous system leukemia(CNSL). Methods The MR and CT imaging data of patients with CNSL between January 2010 and October 2021 were analyzed retrospectively,who underwent head enhanced MR scanning. Results There were 41 patients.MR was positive in 32 cases (78.0%),of which 27 cases (65.9%) were intracranial infiltration,including 14 cases (34.1%) of cranial nerve thickening,13 cases (31.7%) of meningeal enhancement or space occupying,4 cases (9.8%) of brain parenchyma occupation and 1 case (2.4%) of intracerebral ventricular occupation.6 cases (14.6%) of cranial nerve thickening were accompanied by space occupying in areas of cranial nerves. Among the patients with meningeal enhancement or space occupying,there were 7 cases (17.1%) with leptomeningeal enhancement,2 cases (4.9%) with dural enhancement,and 5 cases (12.2%) with dural space occupying. Head CT was performed in 20 cases (48.8%) in the same period,and 9 cases were positive (45%).Among them,7 cases showed intracranial infiltrative lesions,all of which were space occupying lesions with homogeneous slightly high-density shadows. There were 9 cases (33.3%) of missed diagnosis of intracranial infiltrating lesions on MR plain scanning and 7 cases (50.0%) of missed diagnosis on CT. Conclusion The imaging manifestations of CNSL include cranial nerve infiltration,meningeal infiltration,brain parenchyma and intraventricular infiltration and so on. Enhanced MR is of great value in diagnosis,and plain CT has a high detection rate only for space occupying lesions.

Objective To explore the arterial spin labeling (ASL) based on magnetic resonance imaging technology for arterial blood perfusion change pattern and its diagnostic value in patients with migraine without aura. Methods This study prospectively collected 35 migraine patients without aura (MWoA group) and 42 healthy control groups (HC group) who were matched for age,sex and education from January 2020 to December 2022 at Taizhou People's Hospital. The whole brain ASL scan was performed in both groups. Two-sample t test was used to statistically analyze and evaluate the CBF values of each brain region in the two groups. After that,spearman correlation analysis was used to explore the correlation between CBF value of brain area with blood perfusion difference and the course of disease,headache attack frequency,migraine Disability Assessment Scale (MIDAS),Digital Pain Assessment Scale (NRS),headache Impact Measurement-6 score (HIT-6).Receiver operating characteristic (ROC) curve was used to determine whether CBF values in different brain regions could be used to evaluate the diagnostic value of migraine without aura. Results Compared with healthy controls (HC),patients with migraine without aura had lower CBF values in the bilateral middle occipital gyrus (P< 0.05,FEW adjusted).Correlation analysis showed that there was no significant correlation between CBF value of bilateral middle occipital gyrus and duration of disease,headache frequency and migraine related pain scale. ROC analysis showed that the CBF values of the bilateral middle occipital gyrus had good diagnostic value for patients with migraine without aura. When the cut-off values were 1.035 and 1.055,the AUC values were 0.8027 and 0.7980,respectively,both P< 0.001. Conclusion There is abnormal blood perfusion in the brain of patients with migraine without aura. Moreover,the CBF values of these brain regions have ideal diagnostic value for distinguishing migraine patients without aura from normal people.

Objective To explore the efficacy and safety of Pipeline Flex flow diverting device (PED) in the treatment of small and medium-sized intracranial aneurysms of the internal carotid artery. Methods A retrospective analysis was conducted on 18 patients with intracranial aneurysms of the internal carotid artery treated with Pipeline Flex. The Raymond Roy grading was used to evaluate the occlusion of the aneurysms immediately and 6 months after surgery,and the mRS score was used to determine the neurological prognosis at discharge and 6 months after surgery. Results All 18 patients successfully completed the surgery,and 15 patients had no complications. Three cases experienced perioperative complications,two cases were hemorrhagic stroke of which one died,and one case was ischemic stroke. Postoperative VASO-CT showed good adhesion of PED to the wall and patency of the parent artery. There were 2 cases of Raymond Roy grade Ⅰ,5 cases of grade Ⅱ,and 11 cases of grade Ⅲ. At discharge,there were 15 cases with a mRS score of 0,1 case with a score of 1,1 case with a score of 2,and 1 case with a score of 6.The good prognosis rate (mRS score of 0-2) was 94.4%.After 6 months of surgery,DSA reexamination was performed in 14 cases of Raymond Roy grade Ⅰ,2 cases of grade Ⅱ,and 1 case of grade Ⅲ. The mRS score was 0 in 15 cases,1 in 2 cases,6 in 1 case,and the good prognosis rate was 94.4%. Conclusion The surgical success rate for treating small and medium-sized intracranial aneurysms of the internal carotid artery is high,with low complications and acceptable safety.

Objective To evaluate the relationship between motor deficits and crossed cerebellar diaschisis (CCD) after cerebral infarction using 320-slice CT whole-brain perfusion imaging. Methods A total of 115 patients with acute and subacute unilateral supratentorial cerebral infarction confirmed by MRI were enrolled as the experimental group,and 10 normal volunteers were enrolled as the control group. All patients underwent CT whole brain perfusion scan during the stable phase of symptoms,and the motor impairment grade was evaluated in the experimental group. Results CCD occurred in 45 cases in the experimental group,and there was no significant difference in the perfusion value of bilateral cerebellar hemispheres in the control group (P>0.05).In the acute cerebral infarction group,the incidence of CCD in patients with muscle strength score above grade 3 was 21.9%.The incidence of CCD in patients with grade 1-3 was 55%.The incidence of CCD in grade 0 patients was 66.7%.In subacute cerebral infarction group,the incidence of CCD was 35.1% in patients with muscle strength score above grade 3.The incidence of CCD was 50% in the cases between 1 and 3.The incidence of CCD in grade 0 patients was 100%.The muscle strength score of CCD-positive group was lower than that of CCD-negative group in subacute stag(P<0.05) .ΔTTP and ΔMTT in the cerebellum were negatively correlated with the muscle strength score in the CCD-positive group at acute stage,and the reduction rates of ΔTTP,ΔMTT and CBF in the cerebellum in the CCD-positive group at subacute stage were negatively correlated with the muscle strength score. Conclusion 320-slice CT whole-brain perfusion scan can effectively evaluate CCD after cerebral infarction. CCD is more likely to occur in patients with severe motor impairment after cerebral infarction. Cerebellar asymmetry index is related to the degree of motor impairment.

Objective To analyze the X-ray and MRI characteristics of different luminal breast cancers,and to evaluate their predictive value for sentinel lymph node (SLN) metastasis in breast cancer. Methods Clinical data of 88 patients with luminal breast cancer admitted between January 2020 and April 2022 were retrospectively analyzed. Patients were classified into Luminal A and Luminal B according to molecular typing. MRI and molybdenum target X-ray images were performed. Diffusion parameters including apparent diffusion coefficient (ADC),perfusion fraction (f),maximum slope of increase (MSI),and true diffusion coefficient (D) were measured using post-processing techniques. Then receiver operating characteristic curve was plotted to evaluate the effectiveness of each parameter in the differential diagnosis of breast cancers. Results No statistical difference was reported in ADC,f,and MSI values between Luminal A type and Luminal B type (P>0.05),whereas Luminal A type patients yielded a higher D values and higher SER values than Luminal B type patients,with statistical difference (P<0.05).The areas under the curve (AUCs) of MRI-D,MRI-SER,MRI-D/SER,and molybdenum target X-ray in the diagnosis of Luminal A type were 0.869,0.882,0.915,and 0.911,respectively. The AUCs of MRI-D,MRI-SER,MRI-D/SER,and molybdenum target X-ray in the diagnosis of SLN metastasis were 0.885,0.953,0.979,and 0.966. Conclusion MRI-D,MRI-SER,MRI-D/SER and molybdenum target X-ray have high accuracy in differentiating breast cancer subtypes and predicting SLN metastasis.

Objective To explore the value of MRI in the differential diagnosis of benign and malignant non mass like enhancement (NME) lesions. Methods A retrospective analysis was conducted on the imaging and clinical data of 107 female patients (39 cases of benign lesions and 68 cases of malignant lesions) who showed NME on breast MRI images of our hospital and obtained pathological results through puncture or surgery. The distribution types,enhancement methods,ADC values,and time signal intensity curves (TIC) of NME were analyzed and compared between the benign and malignant groups. Multivariate binary Logistic regression analysis was conducted on meaningful influencing factors to screen out risk factors for predicting malignant diseases. Results The results of multivariate binary Logistic regression analysis showed that segment distribution,cluster ring enhancement,type Ⅲ TIC curve,and ADC values were identified as significant factors for predicting malignant breast NME lesions (OR=3.889,95%CI:0.629-24.049;OR=15.367,95%CI:2.602-90.769; OR=7.262,95%CI:1.315-40.099; OR=0.997,95%CI:0.995-0.998,respectively; P<0.05). Conclusion MRI enhanced features and ADC values can effectively improve the differential diagnosis of benign and malignant breast NME lesions.

Objective To explore the prognostic value of proximal lesions based on coronary CT angiography (CCTA) in non-obstructive coronary artery disease (NOCAD). Methods Finally 3359 patients with clinically confirmed or suspected coronary artery disease (CAD) in our hospital from November 2017 to December 2019 were continuous collected. Univariate and Multivariate Cox regression models were used to explore the risk factors of MACE. Results The mean age of patients in the NOCAD group was higher than that in no CAD group,the proportion of men in the NOCAD group was higher than that in no CAD group,and the proportion of smoking,drinking,hypertension and diabetes was higher than that in no CAD group (P<0.001).The mean age and proportion of people with hypertension in NOCAD with proximal lesions group was higher than that in NOCAD without proximal lesions group (P<0.05).There was no significant difference between the two groups in age,gender,smoking,drinking and hypertension (P>0.05).154 patients occurred MACE in NOCAD with proximal lesions group. The incidence rate of MACE was 10.58% and the annual incidence rate of MACE was 2.8% (95% CI:2.4%-3.3%).The HR of NOCAD with proximal lesions group was 1.45 (95% CI:1.09-1.91,P=0.010), when no CAD was used as reference. The HR of NOCAD without proximal lesions group was 1.01 (95% CI:0.67-1.78,P=0.735),when no CAD was used as reference. Conclusion The proximal lesions of NOCAD based on CCTA were 1.45 times higher than those of no CAD. The plaque location based on CCTA may provide additional prognostic value for MACE risk prediction in patients with NOCAD.

Objective To study the MRI findings of primary anterior mediastinal Hodgkin lymphoma (PAMHL) and its value in Ann Arbor clinical staging. Methods A total of 32 patients with PAMHL confirmed by histopathology were enrolled retrospectively. The MRI findings were evaluated independently by two radiologists in double-blinded manner,including tumor size,homogeneity,fibrous septum sign,apparent diffusion coefficient (ADC) value,and so on. Subsequently,we analyzed the inter-observer agreement of these MRI findings and assessed their values in the differential diagnosis of early-stage and advanced-stage PAMHL. Results The size of PAMHL ranged from 5.3 cm to 20.5 cm,with an average length of (9.6±3.7)cm. Twenty-five lesions (78.1%) presented heterogeneous signal intensity; 20 lesions (62.5%) showed fibrous septum sign,18 of which were nodular sclerosis HL; 16 lesions (50.0%) appeared as lymph node fusion; 29 lesions (90.6%) had lymphadenectasis,15 of which showed mass-like lymph nodes with short diameter ≥ 3 cm. Among the 32 PAMHL,tumors invaded lung tissue in 15 cases (46.9%),pericardium in 13 cases (40.6%),blood vessel in 10 cases (31.3%),chest wall in 7 cases (21.9%),and phrenic nerve in 1 case (3.1%).The average ADC value of PAMHL was (0.801±0.112)×10^-3mm²/s. Mass-like lymph node was significantly more seen in advanced-stage than in early-stage PAMHL (P=0.014). Conclusion PAMHL has certain characteristic MRI findings. Mass-like lymph node can be used to distinguish early-stage and advanced-stage PAMHL,which is helpful to guide clinical staging and predict prognosis.

Objective The aim of this study was to investigate the relationship between epicardial adipose tissue volume (EAT) and the risk and load level of frequent premature ventricular complexes (PVC). Methods Retrospective analysis of 80 patients with frequent PVCs and 160 patients without PVCs who underwent computed tomography at the First Affiliated Hospital of Zhengzhou University between October 2020 and December 2021.The evaluation criteria of the 1∶1 propensity to match score (PSM) adjusted analysis were applied. Multifactorial logistic regression was used to analyze the risk factors for the occurrence of frequent PVCs. The relationship between EAT volume and the load level of frequent PVCs was analyzed according to the median number of patients with frequent PVCs divided into a high PVCs burden group and a low PVCs burden group. Results In both the unadjusted and PSM-adjusted cohorts,EAT volume was significantly greater in patients with frequent PVCs than in controls (both P<0.001).Multifactorial logistic regression analysis showed that EAT volume was an independent risk factor for frequent PVCs. (all P<0.001) In patients with frequent PVC,EAT volume was higher in the high PVCs burden group. Conclusion Patients with frequent PVCs have larger EAT volumes compared with controls,and EAT volume is an important independent predictor of freguent PVCs. EAT volume,as an easily measured quantitative index,provides a new insight to differentiate patients with frequent PVCs.

Objective To explore the value of predicting early recurrence (≤ 2 years) after resection of single hepatocellular carcinoma (HCC) based on CT and MRI enhanced portal phase imaging and clinical index prediction models. Methods From April 2016 to February 2020,88 patients with single episode HCC who underwent general surgical resection and were pathologically confirmed at Xuzhou Central Hospital were retrospectively collected. Among them,40 patients with early postoperative recurrence were randomly divided into a training set (n=70) and a validation set (n=18) based on a 4∶1 ratio. Manually segment the entire tumor on CT and MRI enhanced portal phase images to obtain regions of interest and extract imaging features. In the training set,the univariate selection method,least absolute shrinkage and selection operator (LASSO),and Spearman correlation analysis (Spearman) were used to reduce the dimensions of the histological features and select the best feature set. Single factor and multiple factor analysis was used to determine the high-risk clinical factors for early postoperative recurrence in HCC patients. The CT clinical model,MR clinical model,and CT MR clinical model were constructed by combining imaging characteristics and clinical high-risk factors. The prediction value of the model was evaluated using the area under the subject working characteristic (ROC) curve (AUC) and Delong test,and a visual nomogram of the prediction model was generated. Calibration curves were used to evaluate the efficacy of the nomogram,and decision curve analysis (DCA) was used to evaluate the clinical application value of the nomogram. Results Clinical univariate and multivariate analysis showed that aspartate aminotransferase (AST) was an independent risk factor for predicting early recurrence of HCC (P<0.05).Based on CT and MRI enhanced portal phase,a total of 9 CT imaging features and 4 MRI optimal imaging features were selected,and a CT clinical model,an MR clinical model,and a CT MR clinical model were constructed respectively with clinical independent risk factors. The area under the curve (AUC) of the ROC curve in the training set and the validation set were 0.919,0.959,0.971,and 0.844,0.850,and 0.875,respectively. The CT-clinical model was significantly better than the MR-clinical model in predicting early postoperative recurrence of HCC (P<0.05),and the prediction efficiency of the CT-MR-clinical model in the training set was significantly better than the CT-clinical model (P<0.05).DCA showed that the threshold probability of the training set was>30%,and the clinical net income of the CT-MR-clinical model was higher. Conclusion The prediction model based on CT and MRI enhanced portal phase imaging features combined with clinical factors AST (CT MR clinical model) has better predictive efficacy in predicting early postoperative recurrence in HCC patients,and the prediction efficacy of CT clinical model is superior to that of MR clinical model.

Objective The total SPSS diameter (TSD),Total cross-sectional SPSS area (TSA),and Total SPSS volume (TSV) were measured by imaging methods in patients with spontaneous portosystemic shunting (SPSS) ,to compare the diagnostic efficacy for the complications of cirrhosis. Methods A retrospective collection of 107 patients from Shandong University Qilu Hospital Dezhou Hospital from January 2015 to January 2022 was conducted. TSD,TSA and TSV of shunt veins were measured respectively,and the linear relationship was analyzed. Receiver operating characteristic curves (ROC curves) were constructed to predict hepatic encephalopathy,portal vein thrombosis,gastrointestinal bleeding and ascites based on three measurement methods,and the diagnostic efficiency of the three methods was analyzed and compared. Results No linear relationship was found between TSD with both TSA and TSV (B= 0.102,4.176,P> 0.05),but the linear relationship between TSA and TSV existed(B= 9.585,P= 0.015).A total of 46 patients (42.99%) with hepatic encephalopathy were diagnosed accurately by TSD,TSA and TSV (AUC=0.797,0.817,0.810),and there was no statistical significance in the diagnostic efficacy of these three methods(P＞0.05). A total of 39 cases (36.45%) with portal vein thrombosis were diagnosed accurately by TSD,TSA and TSV (AUC=0.751、0.873、0.862),and the diagnostic efficacy of TSA and TSV were higher than that of TSD (P< 0.05).A total of 41 cases (38.32%) with gastrointestinal bleeding were diagnosed inaccurately by TSD(AUC=0.581),but were diagnosed accurately by TSA and TSV(AUC=0.761、0.813),and there was no statistically significant difference in diagnostic efficacy between TSA and TSV (P=0.373).A total of 52 cases (48.60%) with ascites were diagnosed inaccurately by TSD(AUC=0.652),but were diagnosed accurately by TSA and TSV(AUC=0.753、0.878),and the diagnostic efficacy of TSV were higher than that of TSA (P=0.049). Conclusion Imaging measurement had a high predictive value for the occurrence of cirrhosis complications,with TSA and TSV had high efficacy,while TSD had limited efficacy. Considering the convenience of operation,TSA is a more suitable imaging measurement method.

Objective To investigate the feasibility of deep learning image reconstruction (DLIR) algorithm in improving the image quality of thin-slice contrast-enhanced CT(CECT) scan and the lesion conspicuity of pancreatic ductal adenocarcinoma (PDAC) as compared to filtered back projection (FBP) and hybrid model-based adaptive statistical iterative reconstruction (ASIR-V). Methods A retrospective analysis was performed on 44 patients with pathologically confirmed PDAC who underwent preoperative dual-phase CECT scanning. The raw data from portal-venous images were reconstructed using FBP,60% ASIR-V, and DLIR (low, medium, and high strength levels) algorithms at a slice thickness of 0.625 mm. Analysis of variance and Friedman test were used to compare objective (noise in HU, image texture, low-contrast resolution, high-contrast resolution) and subjective (image noise, sharpness, overall image quality and tumor conspicuity) indicators between groups. Results The image texture and low-contrast resolution of DLIR were comparable to FBP but better than 60% ASIR-V, and the noise values of DLIR were lower compared to FBP and 60% ASIR-V (all P＜0.05).All subjective indicators of DLIR were better than or similar to FBP and 60% ASIR-V. With the increase of DLIR strength level, the high-contrast resolution and sharpness scores did not change significantly, the low-contrast resolution and noise decreased, the image texture is blurred, and the overall image quality and tumor conspicuity increased. Conclusion DLIR can significantly improve the image quality of thin-slice enhanced CT images and enhance the lesion conspicuity of PDAC.

Objective To observe the 64-slice spiral CT manifestations of pancreatic cystic lesions and analyze their diagnostic criteria. Methods The clinical data of 70 patients with pancreatic cystic lesions who underwent 64-slice spiral CT examination from March 2020 to March 2023 were collected, of which 33 cases obtained diagnostic results by postoperative pathological examination,8 cases had puncture biopsy pathological diagnosis results, and 29 cases obtained final diagnosis results by clinical follow-up (≥12 months).The 64-slice spiral CT diagnostic results and imaging findings of all lesions were recorded. Results The correct diagnosis rates of pseudocysts, serous cystadenomas, mucinous cystadenomas, and solid-pseudopapillary tumors, intraductal papillary mucinous neoplasm by 64-slice spiral CT were 92.86%,81.82%,80.00%,and 83.33%,80.00%,respectively.There were statistically significant differences in the location, shape, boundary, cyst wall, number of cells, compartments, wall nodules or solid components, peripheral changes, calcification, pancreatic duct dilation, and whether the five cystic lesions were interconnecting with pancreatic ducts (P<0.05).There were significant differences between serous cystadenoma and mucinous cystadenoma in morphology, maximum diameter, cyst wall, cell number and pancreatic duct dilation (P<0.05). Conclusion The internal structure of pancreatic cystic lesions and their relationship with surrounding structures such as pancreatic ducts and blood vessels can be clearly shown by 64-slice spiral CT, which is helpful for accurate preoperative diagnosis in combination with clinical history.

Objective To explore the predictive value of a combined radiomics model and abdominal fat area parameters for microvascular invasion of hepatocellular carcinoma based on CT portal phase. Methods Retrospectively collected 134 cases confirmed by pathology exists Microvascular invasion of Hepatocellular carcinoma scan and enhanced CT images of patients. According to the principle of random stratified sampling,the patients were divided into a training set (n=93) and a testing set (n=41) with a ratio of 7∶3.Firstly,the abdominal fat area parameters of all patients were measured on CT images,and the fat model was constructed by univariate and multivariate Logistic regression. Then,radiomics features of portal venous tumors were extracted by an automated feature extraction algorithm. Spearman correlation analysis and the Least absolute shrinkage and selection operator were used for feature selection. Six kinds of machine learning models were constructed respectively. The better performance model combined with abdominal fat area parameters was used to construct the final combined model. The Area under the curve of the Receiver operating characteristic curve was used to evaluate the predictive efficacy of the model,and the calibration curve was used to verify the calibration capability. The Decision curve analysis is used to analyze and compare the clinical utility of models,and the Delong test assesses the differences in predictive power between models. Results Univariate and multivariate logistic regression indicated that Subcutaneous adipose tissue area and Visceral adipose tissue area parameters were independent risk factors for HCC-MVI,and the AUC values of the fat model constructed by this method were 0.747 (0.648-0.845) and 0.696 (0.536-0.857) in the training and test sets. After a series of feature screening and dimension reduction methods,The remaining 10 radiomics features were used for the construction of six machine learning models,the support vector machine model had better performance. In the training set and the testing set,the AUC values were 0.904 (0.844-0.965) and 0.838 (0.717-0.959),respectively. The joint model was constructed by combining the fat parameters and the radiomics model and visualized by using the nomogram. The performance of the combined model in the training set was 0.925 (0.872-0.978),which was significantly higher than that of the radiomics model (P=0.029) and the fat model (P=0.0054).The performance of the combined model in the testing set was 0.877 (0.772-0.983),which was significantly higher than that of the fat model (P=0.0165),which was higher than that of the radiomics model(P=0.207),the calibration curve showed that the model was well-fitted,and the analysis of the decision curve suggested that the joint model had better clinical practical value than other models. Conclusion The joint model based on radiomics features of the CT portal phase and the abdominal fat area has a high predictive value for HCC-MVI.

Objective To construct and validate a nomogram model based on extracellular volume fraction (ECV) and clinical factors for predicting the expression of p53 in colorectal cancer (CRC). Methods 147 patients with CRC who underwent surgery in our hospital from January 2019 to March 2022 were selected retrospectively. According to pathology,they were divided into two groups:high risk p53 expression group and low risk p53 expression group. Patients were randomly divided into training group (103 cases) and validation group (44 cases).Logistic regression (LR) and lasso regression (LASSO) was used respectively to screen the variables with predictive value,then established several predictive models. The diagnostic efficacy was compared using ROC curves. A nomogram was built basing on the model whose area under the curve was the biggest. The efficiency and clinical benefit of the nomogram model were evaluated by degree of differentiation,calibration and decision curve analysis (DCA). Results The results of multiple Logistic regression showed that sex(OR=2.073,P＜0.05),Maximum diameter of tumor(OR=1.023,P＜0.05) and ECV (OR=0.958,P＜0.05) were independent predictors of p53 high risk expression. LASSO regression screened out 6 potential predictive factors,which was sex,maximum diameter of tumor,ECV,ΔHU tumor,Carbohydrate antigen 199(CA199),neutrophil to lymphocyte ratio (NLR).ROC curves showed that the nomogram model based on LASSO regression has better prediction efficiency(AUC=0.752,95%CI:0.605-0.899,P=0.005).And the calibration curve and DCA curve indicated good calibration and clinical benefit of the model. Conclusion The nomogram model constructed by combining ECV and clinical indicators has good discrimination,calibration and clinical benefit,which can be used to identify CRC patients with p53 high-risk expression non-invasively,and provide reference for clinical diagnosis and treatment.

Objective To investigate the value of spectral CT extracellular volume fraction in preoperative diagnosis of pathological T stage and histological grade of colorectal cancer. Methods The clinical,pathological data and preoperative spectral CT images of 112 patients with pathologically confirmed colorectal adenocarcinoma in the Second and First Affiliated Hospital of Soochow University from June 2022 to December 2022 were retrospectively collected. The quantitative parameters of spectral CT in three enhanced phases of the lesion were measured and analyzed,including CT value,iodine concentration (IC) value,effective atomic number (Eff-Z) value of 40 keV and 100 keV virtual monoenergetic images (VMI) in arterial phase and venous phase,and extracellular volume fraction (ECV) value in delayed phase. The T stage and histologicalgrade of the tumor was obtained by surgical pathology. Independent sample t test or Mann-Whitney U test was used to compare the differences of parameters in different pathological T stages and histological grades. Spearman correlation analysis was used to quantitatively evaluate the correlation between T stages and parameters. The receiver operating characteristic (ROC) curve was used to evaluate the diagnostic effect of parameters with differences between groups. Results The spectral CT quantitative parameters of T3 tumors were higher than those of T1-2 tumors,and the differences were statistically significant (all P< 0.05).Eff-Z_AP,IC_AP,CT 40 keV_AP,CT 100 keV_AP,Eff-Z_VP,IC_VP,CT 40 keV_VP,CT 100 keV_VP,ECV were moderately correlated with T stage (0.4 <r < 0.75),and ECV had the highest correlation with T stage (r=0.602).The Eff-Z_AP,IC_VP,CT 100 keV_VP and ECV values of the high-grade group were higher than those of the low-grade group,and the differences were statistically significant (all P< 0.05).The preoperative diagnostic efficacy of ECV for pathological T stage and histological grade of colorectal cancer was high,with AUC values of 0.885 and 0.738,respectively. Conclusion Spectral CT extracellular volume fraction can be used for preoperative diagnosis of pathological T stage and histological grade of colorectal cancer,and the diagnostic efficiency is better than the conventional quantitative parameters of spectral CT. It can provide a new and non-invasive imaging marker for preoperative evaluation of colorectal cancer,and help to formulate individualized treatment earlier and accurately.

Objective To investigate the value of the radiomics model based on CT images in predicting postoperative recurrence of patients with non-muscle-invasive bladder cancer (NMIBC). Methods The clinicopathological data of 311 patients with NMIBC who underwent transurethral resection of bladder tumor (TURBT) at our institution between January 2014 and December 2017 were collected. The patients were randomly classified into training (n= 218) and validation (n= 93) sets in a 7∶3 ratio. Radiomics features were extracted from the preoperative CT images using the PyRadiomics package. We used the least absolute shrinkage and selection operator (LASSO) regression model to select the optimal combination of the radiomics features and to establish a prognostic classifier,the Rad score. The COX proportional hazards model was used to identify the independent prognostic factors,and a nomogram was then constructed based on these factors. The predictive accuracy,good fit,and clinical utility of the nomogram were evaluated in the training and validation sets,and were compared with the EOTRC recurrence score. Results Using the LASSO model,we identified 5 radiomics features and developed a formula to calculate the Rad score. Multivariate analysis revealed that pT stage,histological grade,and the Rad score were independent predictors of recurrence (P<0.05).A Rad nomogram combining these 3 factors was constructed based on these factors. The nomogram had a good fit as well as higher AUC values (training set:0.762 vs.0.702; validation set:0.828 vs.0.763) and better clinical utility than the EOTRC recurrence score. The Rad nomogram divided the entire cohort into three risk groups,with 5-year recurrence rates of 15.6%,45.2%,and 89.6%,respectively (P<0.001).Moreover,The Rad recurrence grading can differentiate between those with different risks of recurrence in the low, intermediate, and high risk groups of EORTC (P for trend<0.05). Conclusion The CT radiomics model can accurately predict postoperative recurrence of NMIBC patients following TURBT,which may assist in clinical decision process.

Objective To investigate the CT and MRI findings of mixed epithelial and stromal tumor of the kidney (MESTK) in order to improve the understanding of the disease. Methods The CT and MRI features of 7 patients with MESTK confirmed by pathology were analyzed retrospectively. Results There were a total of 8 lesions in 7 patients.3 lesions were cystic and solid masses with more cystic components; The lesions were multicystic structures with or without wall nodules which showed hypointense on T₂WI;3 lesions were cystic and solid masses with more solid components which showed hypointense on T₂WI partly and DWI; The lesion were multiple microcapsules and small cystic structures;2 lesions showed solid lesions with a small amount of fat in the larger one. Conclusion There were some imaging characteristics of MESTK which were helpful for preoperative diagnosis, differential diagnosis and selecting a more suitable surgical plan.

Objective To explore the value of machine learning models based on MRI radiomics features in predicting the degree of meniscus damage. Methods Knee MRI images of 368 patients were retrospectively analyzed. Combined with sagittal and coronal proton density-weighted fat suppression images, the SelectKBest, Minimum redundancy maximum relevance and least absolute shrinkage and selection operator were used to select and reduce the dimension of the radiomics features. Then, based on the optimal features, a variety of machine learning methods are used to build a four-category prediction model, and its performance was evaluated by the receiver operating characteristic (ROC) curve. Results Finally,18 optimal features are obtained. The Macro AUC values of support vector machine, logistic regression, Gaussian process, random forest, quadratic discriminant analysis and Bagging decision tree model were the best, which were 0.876,0.871,0.870,0.869,0.868 and 0.868,respectively.The AUC values of random forest, logistic regression, Bagging decision tree and random forest were the highest (0.948,0.833,0.805,0.902) in the diagnosis of normal meniscus injury, grade 1,grade 2,grade 3 meniscus damage, respectively. Conclusion The machine learning model based on MRI radiomics features for meniscus damage is feasible and has good diagnostic efficiency.

Objective This study examined annulus fibrosus morphology and integrity by measuring the apparent diffusion coefficient (ADC) and fractional anisotropy (FA) using DTI and by discography. Methods This retrospective case–control study enrolled 30 continuous patients with recurrent low back pain (LBP) confirmed by discography and 24 healthy control (HC) volunteers.The ADC and FA values at the posterior margin of the annulus fibrosus were measured using a conventional lumbar MR sequence and axial DTI of the L₃-S₁ levels (3.0-T).The intervertebral discogenic low back pain segment was identified by discography. The ADC and FA values were measured at the posterior margin of the annulus fibrosus of the segment in the experimental group and at the L₃-S₁ level in the control group. Differences in ADC,FA,λ1,λ2 and λ3 values were compared between the two groups. Results Thirty participants (58 ± 6 years,16 women) were in the LBP group (case),twenty-four (54±6 years,12 women) were in the HC group (control).No significant differences were found in demographic parameters or the areas of region of interest (ROI) measured manually between the two groups. The ADC value of ROI in the experimental group was higher than that in the control group (LBP:9.84 ± 4.01; HC:6.10 ± 3.17; P<0.001),while the FA value was lower than that in the control group (LBP:0.231 ± 0.150; HC:0.341 ± 0.201; P<0.007). Conclusion 3.0-T magnetic resonance DTI was useful in the clinical assessment of discogenic low back pain.

Objective To investigate the imaging and clinical characteristics of clear cell sarcoma of the Kidney (CCSK) and malignant rhabdoid tumor of the kidney in children. Methods We conducted a retrospective analysis the case data of the 4 children with CCSK and 4 children with MRTK confirmed by pathologically.All 8 cases were examined with plain and enhanced CT scans preoperative. Results 4 cases of CCSK,1 was located in the left kidney and 3 in right kidney.4 cases of MRTK,3 were located in the left kidney,1 was in right kidney. The 8 lesions were all circular. The tumors were all large,and the maximum diameter of the CCSK lesions ranged from 34 mm to 149 mm,with an average of (73±35)mm. The maximum diameter of MRTK lesions ranged from 41 mm to 81 mm,with an average of (62±15)mm.4 cases of CCSK mainly showed residual kidney morphology,kidney tissue around the lesion. The lesion was located in the renal pelvis with deformation and enlargement in 3 cases,and broke through the renal pelvis and grew outward in 1 case. The main findings showed equal or slightly low density mass,but uneven slightly high density mass with more bleeding no Non-enhanced CT. Showed mild to moderate uneven progressive enhancement after contrast—enhanced scanning. Necrosis/cystic change and hemorrhage were observed in all 4 cases. Subcapsular effusion/blood was observed in 1 case,namely “crescent sign”; No calcification or bone destruction was observed. Lung and lymph node metastases in 1 case. Among the 4 cases of MRTK,there is dual kidney disappeared in 3 cases and only 1 existed. Residual renal pelvis were pressed and pushed. The lesions were heterogeneous solid mass,mostly with cystic degeneration,and the boundary between cystic and solid was not clear,Characterized by changing gradually on Non-enhanced CT,manifested as “melting ice sign”. Showed mild to moderate uneven progressive enhancement after contrast—enhanced scanning. There was necrosis/cystic change in all the 4 cases.3 cases showed high density bleeding shadow; Calcification was found in 2 cases; “Crescent sign” appeared in 4 cases; No bone destruction was observed. Tumor thrombus was found in 1 case of brain metastasis,1 case of lung and lymph node metastasis,and 1 case of inferior vena cava and left renal vein. Conclusion The incidence of CCSK and MRTK in children is low,the CT manifestations of them have many similarities with other common renal malignancies in children,and the lack of understanding of the signs are the main reasons for misdiagnosis. However, a comprehensive understanding of the imaging manifestations of CCSK and MRTK and misdiagnosis analysis can improve the diagnostic accuracy for the disease.

Objective To investigate the value of 3D-FIESTA sequence of MRI in fetal spinal malformation. Methods MRI spinal images of 58 fetuses were analyzed retrospectively. All pregnant women were diagnosed with suspected spinal abnormalities by prenatal ultrasound before MRI acquisition. Grade the image quality and diagnostic value of 2D-FIESTA and 3D-FIESTA sequences,and compare the image quality of the two groups. Results A total of 58 fetal spinal images were obtained by scanning. There were 19 cases of hemivertebrae,7 cases of butterfly vertebrae,5 cases of fusion vertebrae,3 cases of coronary fissure,14 cases of small ossification center of vertebral,3 cases of narrow intervertebral space,4 cases of uneven ossification of vertebrae,and 14 cases of negative results. The excellent,medium and poor grading of 2D images accounted for 31.03%,60.34% and 8.63% respectively,while the 3D images accounted for 81.03%,15.52% and 3.45% respectively. The quality of 3D images is higher than that of 2D images,and the difference is statistically significant (P<0.05). Conclusion 3D-FIESTA sequence can better display fetal spinal malformation than 2D.When 2D sequence scanning is not satisfactory,3D sequence can be added to improve the detection rate of fetal spinal malformation.

Objective To analyze the CT and MR manifestations of meningioma in children,to improve the diagnostic accuracy of this disease. Methods Retrospectively analyzed the clinical and imaging data of 14 cases with meningioma confirmed by surgery and pathology,and reviewed relevant literatures. Results There were 9 males and 5 females,aged from 10 months to 18 years,with an average age of 98.35 months. The size of the mass was 16 mm×14 mm×15 mm~63 mm×79 mm×90 mm.9 cases underwent CT examination,and 14 cases underwent MR examination. The tumors located in the sellar area (1 case),parasellar area (1 case),pineal area (1 case),orbit (1 case),lateral ventricle (1 case),cerebellopontine area (1 case),parafalcine (2 cases),parietal-frontal-temporal region (5 cases),respectively,with diffuse and multiple distributing in meninges and cerebral falx in 1 case.13 cases with single lesion and 1 case with multiple lesions.2 cases were accompanied with neurofibromatosis. There were 7 cases of WHO grade Ⅰ,4 cases of WHO grade Ⅱ,and 3 cases of WHO grade Ⅲ. Solid part showed iso-density or slightly hyper-density on CT,iso-intensity mainly on T₁WI and T₂WI,and iso-or slightly hyperintensity on DWI. Lesions showed slightly hypo-intensity on T₂WI in 3 cases. Lesions showed obvious enhancement and most of them were heterogeneous. Most of the cases with intracranial lesions were accompanied by edema. Rate of misdiagnosis was 50% (7/14). Conclusion Meningiomas in children are large and most of them show cystic area. With high incidence of the high-grade meningioma and different distribution comparing with adults,meningiomas in children could be misdiagnosed easily. Combing with clinical information,some specific imaging manifestations can help improve the diagnostic accuracy of this disease.