http://med-sci.cn/cgzcx/en/contents_list.asp?issue=202311 zh-cn An RSS feed for Chinese Journal of Magnetic Resonance Imaging http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.001 ObjectiveTo investigate the effects of smoking and drinking on brain spontaneous activity and the neurobiological mechanism of comorbidity between smoking and drinking. Materials and MethodsBased on the amplitude of low-frequency fluctuation (ALFF) method, a 2×2 factorial design was employed to investigate the effects of alcohol on ALFF in individuals categorized into four groups: alcohol-consuming smokers (n=35) and alcohol-consuming non-smokers (n=27), as well as non-alcohol-consuming smokers (n=21) and non-alcohol-consuming non-smokers (n=25). Resting-state functional magnetic resonance imaging (fMRI) scans were performed to acquire brain data, and ALFF values were calculated for each group. Subsequently, an analysis of variance (ANOVA) was conducted to assess the differences in ALFF values among the four groups. To further explore specific group differences, post hoc tests (Bonferroni correction) were conducted based on the ANOVA results. The study aims to gain insights into the impact of alcohol consumption on ALFF in the context of nicotine dependence, contributing to a deeper understanding of neural activity and its modulation in addiction. ResultsCompared with healthy controls, the smoking group had higher ALFF values in the bilateral precuneus and right cuneus (t=3.212, P=0.001), while the drinking group had higher ALFF values in the left cerebellar hemisphere (t=3.422, P=0.001), with P＜0.005 for voxel levels and P＜0.05 for mass levels, based on GRF correction. Smoking and alcohol consumption jointly affected brain activity in the left posterior cerebellar lobe, but with opposite effects. ALFF values in the left posterior cerebellar lobe were negatively correlated with smoking age, pack-years and drinking dependence scale scores (n=35; r=-0.367, P=0.025; r=-0.267, P=0.033; r=-0.293, P=0.026). ConclusionsSmoking and alcohol consumption both affect spontaneous brain activity, and their interaction occurs in the left posterior cerebellar lobe. Smoking and drinking produce a novel antagonistic interaction. This suggests that we need to control for alcohol consumption as a variable when studying spontaneous brain activity in smokers. ]]> Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.002 ObjectiveSurface-based morphometry (SBM) and voxel-based morphometry (VBM) were used to measure the whole brain morphology of frontal lobe epilepsy in children and adolescents with negative MRI. The morphological characteristics of brain microstructure changes in children and adolescents with epilepsy were discussed. The correlation between the brain regions with the change of brain microstructure and the course of disease and intelligence quotient (IQ) was analyzed. Materials and MethodsChildren and adolescents with epilepsy were collected from the Department of Neurology and Pediatrics of our hospital. The diagnosis of children with epilepsy should meet the diagnostic criteria established by the International League Against Epilepsy (ILAE) in 2017. Patients with frontal lobe epilepsy were selected as case group according to the medical history, electroencephalogram analysis and clinical symptoms of the children. Healthy children were recruited as control group. Clinical information of the case group and control group should be recorded in detail. Two groups were tested with MRI to obtain high-resolution 3D-T1WI structural image data and T2-fluid attenuated inversion recovery (T2-FLAIR) data. The case group underwent an intelligence test using the China-Wechsler Intelligence Scale for Children within 3 days of examination. SBM and VBM used FreeSurfer software and VBM software. SPSS 23.0 was used to analyze the correlation between brain areas and disease duration and IQ in children with MRI negative frontal lobe epilepsy, and compare the similarities and differences of abnormal brain areas obtained by SBM and VBM. ResultsThere was no significant difference in gender, age and education between the two groups (P＞0.05), compared with the control group, the results of SBM showed abnormal brain areas in the case group: the frontal pole, superior frontal gyrus, posterior central gyrus, middle frontal gyrus, inferior frontal gyrus and bilateral superior temporal gyrus in the left cerebral cortex decreased (P＜0.05), the precuneus and precuneus and bilateral superior temporal gyrus in the superior, middle and inferior gyrus of the left and right cerebral cortex decreased reduced (P＜0.05); the results of VBM analysis showed reduced gray matter volume in the left inferior temporal gyrus, right screen nucleus and middle frontal gyrus and bilateral posterior central gyrus in children and adolescents with epilepsy (P＜0.05), there was no significant correlation between abnormal brain areas and disease course and IQ (P＜0.05). ConclusionsSBM and VBM analysis methods can both detect brain regions with abnormal brain morphology in children and adolescents with frontal lobe epilepsy, and SBM can extract more abnormal brain regions than VBM analysis; morphological changes of middle frontal gyrus and posterior central gyrus were observed in both analysis methods; the abnormal changes in the brain regions of children and adolescents with frontal lobe epilepsy mainly involve the frontal lobe and peripheral brain regions. ]]> Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.003 ObjectiveTo utilize the resting-state functional magnetic resonance imaging technology to explore the changes of voxel-mirrored homotopic connectivity (VMHC) in patients with severe obstructive sleep apnea (OSA). Materials and MethodsA total of fifty patients with OSA and fifty healthy controls, matched in terms of age, sex, and education, were included in this study. The data of clinical situations, neuropsychological scale assessment and brain magnetic resonance imaging of all participants were further collected. VMHC and the seed-based functional connectivity (FC) were calculated and compared between these two groups. Additionally, Pearson correlation analysis was conducted to examine the relationship between the significant brain areas for VMHC and FC, and the clinical variables and neuropsychological scale scores. ResultsCompared to control group, patients with severe OSA exhibited lower cognitive scores, and higher depression and anxiety scores. In patients with OSA, the VMHC of the bilateral calcarine and the bilateral superior parietal gyrus was significantly decreased. The FC was found to be abnormal in the bilateral lingual gyrus, right middle occipital gyrus, and left middle temporal gyrus (gaussian random field correction, voxel level P<0.001, cluster level P＜0.05). The VMHC value of bilateral calcarine gyrus in severe OSA patients showed a negative correlation with the apnea hypopnea index (r=-0.31, P=0.027), and a positive correlation with mean blood oxygen saturation (SaO₂) (r=0.30, P=0.033). Additionally, the VMHC value of the bilateral superior parietal gyrus was positively correlated with the mean and minimum SaO₂ (r=0.29, P=0.039; r=0.31, P=0.028). Furthermore the FC value between the left superior parietal gyrus and the angle gyrus was positively correlated with the mean SaO₂ (r=0.29, P=0.041). ConclusionsCognitive function impairment and potential risk of depression and anxiety are observed in severe OSA patients. The alteration of interhemispheric coordination and abnormal FC of the bilateral calcarine and the bilateral superior parietal gyrus may be significant neuropathological mechanisms contributing to cerebral impairments in patients with OSA. ]]> Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.004 ObjectiveTo investigate the clinical utility of compressed sensing (CS) combined with echo-planar-imaging (EPI) arterial spin labeling (EPIC-ASL) in improving cerebral perfusion imaging, by compared with traditional EPI-ASL images. Materials and MethodsWe prospectively enrolled 26 patients with acute cerebral infarction (ACI) and 32 healthy volunteers. They were scanned both EPI-ASL and EPIC-ASL program. Two radiologists independently evaluated the imaging quality of the white matter, gray matter, basal ganglia, brainstem and cerebellum acquired by EPIC-ASL and EPI-ASL, respectively. Then, the signal to noise ratio (SNR) and gray/white matter contrast noise ratio (CNR) were caculated in each anatomical region same as above. For patients with ACI, the boundary of infarction, SNR, CNR _{infarction/white matter}, and relative blood flow (rCBF) were analyzed. The paired-sample t-test, Wilcoxon rank sum test, and Mann-Whitney U test were used in appropriate to compare the image quality between the two groups. ResultsEPIC-ASL performed better than EPI-ASL for displaying white matter, gray matter, basal ganglia, brainstem and cerebellum, (all P＜0.001). In each anatomical location, the SNR and CNR_{gray matter/white matter} of EPIC-ASL were all considerably higher than those of EPI-ASL (all P＜0.001). Compared with EPI-ASL, EPIC-ASL could depict the boundaries of infarct more accurately (P＜0.001) and showed higher SNR and CNR_{infarction/white Matter} values (P＜0.001 and P＜0.032, respectively). There was no significant difference in the evaluation of rCBF between the two techniques (P=0.851). ConclusionsWithin the same scanning time, EPIC-ASL can improve the in-plane resolution and image quality compared with traditional EPI-ASL. EPIC-ASL shows better performance for the visualization of infarction, and can accurately assess cerebral perfusion, which will be benefit for patients with ACI. ]]> Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.005 ObjectiveTo explore the efficacy of multimodal imaging radiomics models of different tumor microenvironment subregions in predicting the methylation status of the O⁶-methylguanine-DNA methyltransferase (MGMT) promoter in glioblastoma before surgery. Materials and MethodsA retrospective analysis was conducted on preoperative MRI images, clinical, and genetic information of 600 glioblastoma patients from Taihe Hospital, Hubei University of Medicine, University of Pennsylvania, and University of California, San Francisco. The preprocessed images were automatically segmented to obtain three subregions of the tumor microenvironment. From the preoperative MRI images [contrast enhanced T1-weighted imaging (T1WI-CE), T2 fluid attenuation inversion recovery (T2-FLAIR) sequence, and diffusion tensor imaging (DTI) fractional anisotropy (FA) maps], 2 153 radiomics features were extracted from three habitat subregions, including enhanced region, necrotic region and edema region. Feature selection was performed using correlation analysis, minimum redundancy maximum relevance (MRMR), and Boruta algorithm, and the XGBoost algorithm was used to build classification model. The diagnostic performance of the models was evaluated using receiver operating characteristic (ROC) curves, area under the curve (AUC), accuracy, sensitivity, specificity, and DeLong test for model comparison. ResultsThere were no statistically significant differences in the intergroup comparisons of clinical features between the two subtypes in the training and testing sets (P＞0.05). The multimodal imaging radiomics model for the enhanced region had AUCs of 0.842 and 0.935 in the training and validation sets, respectively. Ten features from the multimodal habitat subregions were obtained after feature selection. The AUC of the imaging omics model in the multimodal habitat subregion was 0.874 and 0.899 on the training and test sets, respectively. ConclusionsThe preoperative MRI radiomics models can predict the MGMT promoter methylation status in glioblastoma patients, and the multimodal combination models showed more robust diagnostic performance. The study of tumor microenvironment subregions provides important clinical utility for accurate molecular subtyping, decision-making for temozolomide (TMZ) use, and survival prediction in glioblastoma patients. ]]> Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.006 ObjectiveTo evaluate the association between myocardial fibrosis by late gadolinium enhancement (LGE) on cardiac magnetic resonance (CMR) imaging and adverse outcomes in patients with hypertrophic cardiomyopathy (HCM). Materials and MethodsWe searched databases including PUBMED, EMBASE, the Cochrane Library, CBM, CNKI and Wanfang database from the time of database establishment to August 2023, and selected observational cohort studies evaluating the association between myocardial fibrosis by CMR-LGE and adverse events in HCM patients, and then screened according to inclusion and exclusion criteria. A Meta-analysis was performed to combine and establish hazard ratio (HR) of results. All statistical analyses were performed using STATA 15.0. ResultsData extracted from 17 studies including 6576 patients with HCM were reviewed. Totally, the presence of LGE had increased risk of combined adverse end points {HR=2.34 [95% (confidence interval, CI): 1.55-3.53], P＜0.01} and heart failure (HF) [HR=2.32 (95% CI: 1.37-3.93), P＜0.01] in HCM patients. And myocardial LGE positive was not significantly associated with increased risk of cardiac death [HR=1.86 (95% CI: 0.80-4.30), P=0.15] and sudden cardiac death (SCD) [HR=1.98 (95% CI: 0.64-6.15), P=0.24] in HCM patients. The amount of LGE (per 10% increase) was associated with increased risk for combined adverse end points [HR=1.53 (95% CI: 1.35-1.74), P＜0.01], all-cause death [HR=1.42 (95% CI: 1.22-1.65), P＜0.01], cardiac death [HR=1.55 (95% CI: 1.34-1.80), P＜0.01], SCD [HR=1.51 (95% CI: 1.28-1.79), P＜0.01] and HF [HR=1.51 (95% CI: 1.26-1.80), P＜0.01] events in HCM patients. ConclusionsMyocardial LGE is associated with increased risk of adverse outcomes in HCM patients and can be used as an independent predictor of prognosis in HCM patients. ]]> Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.007 ObjectiveTo evaluate the diagnostic value of quantitative T1 mapping in cardiac magnetic resonance imaging for myocardial tissue characteristics in patients with pulmonary hypertention (PH) by Meta-analysis. Materials and MethodsRelevant literatures were retrieved from Pubmed, Web of Science, the Cochrane Library, CNKI, Wanfang Database and VIP Database from the establishment of the database to December 31, 2022. Two researchers independently screened literatures, extracted data and evaluated the quality of the included literatures according to the inclusion and exclusion criteria. Meta-analysis was performed using Revman 5.4 software. ResultsA total of 8 literatures were included, including 613 PH patients and 150 healthy controls. The results of Meta-analysis showed that compared with the healthy control group, the initial T1 value of PH patients was significantly increased [odds ratio (OR)=123.22, 95% CI (112.92-133.52), P＜0.001]. Myocardial damage was obvious in inferior right ventricle insertion point (IRVIP) in PH group [OR=-82.94, 95% CI (-92.43--73.45), P＜0.001]. ConclusionsMR T1 mapping technique can evaluate the myocardial tissue characteristics of PH patients noninvasculatively, which has high clinical application value. Close attention should be paid to the myocardium of IRVIP. ]]> Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.008 ObjectiveTo explore the predictive value of T1WI-enhanced radiomics model for non-small cell lung cancer brain metastases in predicting epithelial growth factor receptor (EGFR) mutation status in non-small cell lung cancer. Materials and MethodsThe imaging data of cranial magnetic resonance non-contrast scan + contrast examination of 97 patients with non-small cell lung cancer brain metastases before treatment were retrospectively analyzed (50 EGFR mutant and 47 EGFR wild type), and randomly grouped into training group and test group according to 8 : 2. The radiomics features were extracted from the T1WI-enhanced transverse, coronal and sagittal positions, and the dimensionality reduction and screening of the radiomics features were successively carried out by VarianceThreshold, SelectKBest and least absolute shrinkage and selection operator (LASSO), and the support vector machines (SVM) and logistic regression (LR) were used for classifier modeling, and cross-validation by 5-fold method, finally the performance of the prediction model was evaluated in the test group, the receiver operating characteristic (ROC) curve of the training group and the test group was drawn to evaluate the prediction efficiency, and the difference between the models is compared by DeLong test. ResultsThe AUCs of T1WI enhanced transverse, coronal and sagittal radiomics models reached 0.64, 0.68 and 0.80, respectively. The AUC of the combined sequence model test group can reach 0.84, among which the LR classifier has the best prediction efficiency, the AUC, sensitivity, specificity and accuracy of the training group are 0.86, 74%, 75% and 76%, respectively, and the AUC, sensitivity, specificity and accuracy of the test group are 0.84, 80%, 78% and 80%, respectively, and the DeLong test between the models has no significant significance (P＞0.05). ConclusionsRadiomics model based on T1WI enhanced transverse, coronal and sagittal positions can predict EGFR mutation status, and the LR classifier model combined with sequence has the best prediction efficiency, which is helpful to guide the rational selection of targeted drug therapy and individualized precision medicine in clinical practice. ]]> Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.009 ObjectiveTo investigate the diagnostic value of multiparametric MRI image features and related parameters for human epidermal growth factor receptor 2 (HER-2) low expression breast cancer, to predict the expression status of prognostic factors within the tumor tissue, and to summarize the characteristics of MRI signs of HER-2 low expression breast cancer. Materials and MethodsFifty-two patients with HER-2 low-expression breast cancer who received treatment from January 1, 2014 to April 12, 2022 were selected as subjects for retrospective study. The MRI sign characteristics of HER-2 low expression breast cancer were analyzed. The results of immunohistochemical staining examination were also used as the gold standard, and the results were analyzed according to the prognostic factors [estrogen receptor (ER), progesterone receptor (PR), marker of proliferation Ki-67 (Ki-67), and the positive/positive ratio of HER-2] positive/negative expression were grouped four times. The intra-class correlation coefficient (ICC) was used to evaluate the consistency of the results of the basic characterization of MRI images between different physicians. Depending on the type of raw data, the differences of each parameter between the groups were compared and analyzed using one-way analysis (independent samples t-test, Fisher^,s exact probability method, Mann-Whitney U-test, etc.), respectively, and a joint diagnostic model was established by using biclassified logistic regression analysis to compare the differences of the basic clinicopathological and MRI signs among the four groups, which led to the conclusion that the differences of the basic clinicopathological and MRI signs between the four groups of HER-2 low The common MRI signs of HER-2 low expression breast cancer were characterized, and the correlation between clinical imaging features and prognostic factors was explored. ResultsThe results of basic characterization of MRI images were in good agreement between different physicians (ICC range 0.883-0.972). HER-2 low-expression breast cancer clinicopathologically and on MRI mostly showed age 29-74 (51.10±10.67) years old, pathologic type: non-specific invasive breast cancer accounted for the majority (50/52, 96.2%); histologic grading was predominantly Ⅱ-Ⅲ (47/54, 87.0%); MRI signs mostly showed a single lesion (41/52, 78.8%), burr edges (33/52, 63.5%), lobular signs (30/52, 61.5%), internal inhomogeneous enhancement (36/52, 69.2%), time intensity curve (TIC) type Ⅲ predominantly (46/52, 88.5%), and almost no or small amount of background parenchymal The mean value of apparent diffusion coefficient (ADC) was about (0.767±0.143)×10^-3 mm²/s, and the range of values was (0.512-1.200)×10^-3 mm²/s. The prognostic value of clinicopathological and MRI features on prognostic factors: ER, PR, and TIC. Predictive value: In the ER and PR positive group, burr sign was mostly seen at the edge of the mass (ER and PR positive: 72.1%, P=0.008), internal enhancement was more heterogeneous (ER positive: 76.7%, P=0.030, PR positive: 79.1%, P=0.003), and histologic grading was lower (ER positive: P=0.008, PR positive: P=0.003). In the ER and PR negative group, there were more blurred margins (ER and PR negative: 55.6%, P=0.008), and internal enhancement was predominantly circumferential (ER negative: 66.7%, P=0.030, PR negative: 77.8%, P=0.003). In the Ki-67 positive group, the morphology of the mass was more lobular (68.2%, P=0.034) compared to the negative group, with a higher histologic grading was higher (P=0.003). Two-category logistic regression analysis suggested that histologic grading was an independent correlate for predicting the expression of ER, PR, and Ki-67 in HER-2 low-expressing breast cancers (P=0.032, P=0.022, P=0.003), and internal enhancement features were an independent correlate for predicting the expression of ER and PR in HER-2 low-expressing breast cancers (P=0.041, P=0.014). ConclusionsThe clinicopathologic and MRI features of HER-2 low-expression breast cancer have certain specificity, and multiparametric MRI is valuable for the early clinical diagnosis of HER-2 low-expression breast cancer and the prediction of the expression status of its related prognostic factors. ]]> Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.010 ObjectiveTo explore the application value of virtual magnetic resonance elastography (vMRE) based on diffusion weighted imaging (DWI) in focal liver lesion (FLL). Materials and MethodsA retrospective analysis was made of 224 patients with FLL who underwent upper abdomen MRI in the First Affiliated Hospital of Xi^,an Jiaotong University from November 2021 to June 2022. All patients upper abdominal MRI examination sequences including outine fat suppression T2WI (FS-T2WI) and multi b-value DWI. The study subjects were divided into five groups based on pathological or clinical diagnosis: hepatocellular carcinoma, metastatic tumor, cholangiocarcinoma, hemangioma, and liver cyst. Compare the measured values between groups for statistical differences using t-tests or Mann Whitney U-tests, and draw receiver operating characteristic (ROC) curves to evaluate the diagnostic efficacy of vMRE. ResultThere was a statistically significant difference in the stiffness value of vMRE in distinguishing benign and malignant lesions (Z=-12.309, P＜0.01), and there was a statistically significant difference in the stiffness value between differentiating hemangiomas and malignant lesions (Z=-6.733, P＜0.01). The sensitivity, specificity and area under the curve (AUC) of virtual elastography were 92.2%, 96.7% and 0.981 respectively in differentiating benign and malignant FLL. The sensitivity of virtual elastography in differentiating hemangioma from malignant lesions was 92.2%, the specificity was 88.0%, and the AUC was 0.926. There were statistical differences between hemangioma and hepatocellular carcinoma (Z=-6.232, P＜0.01), metastatic tumor (Z=-5.975, P＜0.01), and cholangiocarcinoma groups (Z=-4.313, P＜0.01). In malignant lesions, there was no significant statistical difference between the hepatocellular carcinoma group and the metastatic tumor group (K=1.231, P＞0.05), and the hepatocellular carcinoma group and the cholangiocarcinoma group (K=-1.403, P＞0.05); There was a statistical difference between the metastatic tumor group and the cholangiocarcinoma group (K=-2.062, P＜0.05). ConclusionsvMRE based on DWI is a non-invasive indicator that reflects the stiffness of tissues. It is helpful for the differential diagnosis of common benign and malignant liver lesions, and may provide a new indicator for the differential diagnosis of atypical hemangiomas and malignant lesions in clinical practice. ]]> Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.011 ObjectiveTo compare clinical value of magnetic resonance cholangiopancreatography (MRCP) using three-dimensional respiratory-triggered turbo spin-echo (3D RT-TSE) sequence and three-dimensional breath-hold gradient and spin-echo (3D BH-GRASE) sequence in the diagnosis of extrahepatic cholelithiasis. Materials and MethodsA retrospective analysis was made on 74 patients who underwent MRCP due to clinical suspicion of cholelithiasis at the Affiliated Zhangjiagang Hospital of Soochow University from November 2017 to November 2022. Both 3D RT-TSE and 3D BH-GRASE sequences were employed for MRCP. Images of the two sequences were evaluated by three radiologists with 3, 6 and 9 years of experience in abdominal MRI, respectively. Three radiologists independently evaluated the overall image quality of the acquired images and assessed the diagnostic quality of images suitable for diagnosis. The Wilcoxon signed-rank test was utilized to compare the overall image quality of the two sequences. The McNemar^,s test was employed to assess the differences in results obtained by the readers using the two imaging methods. ResultsThe overall quality scores of 3D BH-GRASE MRCP images were superior to those of 3D RT-TSE MRCP images (Z=-7.286, P＜0.001). In the diagnosis of gallbladder stones, the sensitivity, specificity and accuracy of 3D BH-GRASE MRCP were 92.3%, 89.7% and 91.2%, respectively, while for 3D RT-TSE MRCP, these values were 66.7%, 86.2% and 75.0%, respectively. The difference of sensitivity, and accuracy between the two groups was statistically significant (P≤0.002), but the difference of specificity was statistically insignificant (P=0.317). In the diagnosis of extrahepatic bile duct stones, the sensitivity, specificity and accuracy of 3D BH-GRASE MRCP were 72.7%, 91.3% and 85.3% respectively, whereas for 3D RT-TSE MRCP, these values were 68.2%, 91.3% and 83.3%, respectively. There was no significant difference between them (P＞0.05). Conclusions3D BH-GRASE MRCP is superior to 3D RT-TSE MRCP in the detection of extrahepatic cholelithiasis. Reasonable optimization of MRCP sequence could enhance the efficiency of examination and diagnostic efficacy. ]]> Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.012 ObjectiveTo explore the prognostic factors of acute pancreatitis (AP) and build a nomogram model based on clinical manifestations and MRI features to predict the prognosis of peripancreatic collections (PPCs) in early stages of AP. Materials and methodsWe conducted a retrospective study involving 102 patients with peripancreatic collections of AP admitted to our hospital from January 2016 to February 2023. Patients were categorized into two groups based on clinical and imaging follow-up outcomes: good prognosis and poor prognosis. A multifactorial logistic regression analysis was performed to establish a nomogram based on clinical manifestations and MRI features for early prognosis prediction. The nomogram^,s performance was evaluated by calculating the area under the curve (AUC) of receiver operating characteristic (ROC), bootstrap resampling, and visual inspection of the calibration curve. Additionally, decision curve analysis (DCA) was used to assess the nomogram^,s clinical utility. ResultsOf the 102 AP patients with peripancreatic collections, 77 had a good prognosis, while 25 had a poor prognosis. Compared to the good prognosis group, indicators associated with poor prognosis, such as complications of AP, Extrapancreatic Inflammation on MRI (EPIM), Bedside Index for Severity in AP (BISAP), the number of peripancreatic spaces related to PPCs, MR Severity Index (MRSI), the maximum dimensional area of PPCs, classification of PPCs, the diffusion range grading of PPCs involving sub-peritoneal spaces, and involvement of the abdominal wall and peripancreatic vessels, showed a significant increase in proportion (P＜0.05). Multifactorial prognostic analysis revealed that MRSI, EPIM, maximal dimensional area of PPCs, and BISAP score were independent prognostic factors for early prognosis prediction of PPCs in AP. The nomogram demonstrated excellent performance, with an AUC of 0.946 (95% CI: 0.905-0.988) and excellent agreement between predicted and observed probabilities, as indicated by calibration curves. The DCA curve further confirmed its clinical utility. ConclusionsA nomogram model based on clinical manifestations and MRI features can effectively predict the early prognosis of PPCs in AP. ]]> Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.013 Objective3.0 T MRI data has clinical value in evaluating the efficacy of neoadjuvant therapy for locally advanced rectal cancer (LARC), but the comparison between multiple machine learning models has not been explored. We will compare the efficacy of four commonly used machine learning methods in evaluating the clinical value of neoadjuvant chemoradiotherapy (nCRT) for LARC. Materials and MethodsA total of 160 LARC patients who were diagnosed and confirmed by pathological examination at the Second Affiliated Hospital of Harbin Medical University from September 2021 to January 2023, underwent nCRT. They were divided into a training set and a validation set in an 8∶2 ratio. Establish four classifier models: support vector machine (SVM), naive Bayes (NB), convolutional neural networks (CNN) and neural network (NN), and use DeLong test to compare the differences in receiver operating characteristic (ROC) curves. Evaluate and compare the diagnostic performance of four classifiers. ResultsThere was no statistically significant difference in age and gender between the two groups of patients (P＞0.05). Nine features related to treatment efficacy grouping were obtained through least absolute shrinkage and selection operator (LASSO), and there were differences between pathological complete response (pCR) non-pathological complete response (non-pCR) groups, but the differences were not statistically significant (P＞0.05). The area under the ROC curve of SVM in the training set is 0.9150, which indicates the most significant evaluation of the efficacy of nCRT and chemotherapy. ConclusionsBased on the texture features of high-resolution T2WI MRI, SVM, NB, NN, and CNN classifier models can be used to evaluate the effectiveness of colorectal cancer nCRT treatment. SVM classifier models have the best diagnostic performance, and imaging omics based on high-resolution T2WI can evaluate the effectiveness of nCRT treatment in LARC patients. ]]> Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.014 ObjectiveTo explore the value of deep learning (DL) imageomics based on MRI axial high-resolution T2WI images in predicting T2 and T3 stages of rectal cancer before surgery. Materials and MethodsRetrospective analysis of the complete data of 361 patients with T2 and T3 stage rectal cancer confirmed by postoperative pathology at the First Affiliated Hospital of Wannan Medical College (Yijishan Hospital) from January 2018 to December 2022. Among them, there were 100 cases in T2 phase and 261 cases in T3 phase. Patients were randomly divided into a training set (n=262) and a testing set (n=99) using stratified sampling at 7∶3. Univariate and multivariate logistic regression analysis was used to screen independent risk factors for clinical imaging features. The ResNet-18 model was employed as the foundational model for DL feature extraction. Subsequently, twelve machine learning models were developed by incorporating clinical imaging features, hand‑crafted radiomi features, DL features, and their combined features. The support vector machine (SVM), K-nearest neighbor (KNN), and extreme gradient enhancement machine (XGBoost) algorithms were utilized for constructing these models. The diagnostic performance of each model was assessed by calculating the area under the curve (AUC) of the subject. Finally, the model with the highest performance was identified as the optimal output model. ResultsThe results of both univariate and multivariate logistic regression analysis indicate that carbohydrate antigen (CA19-9) [95% confidence interval (CI): 1.150-1.820, P=0.002] and tumor length (LD) (95% CI: 1.159-22.584, P=0.031) were independent risk factors for predicting T2 and T3 stage rectal cancer based on clinical imaging features. Among all the models constructed, the performance of combined feature model was higher than that of individual feature model, and the training set XGBoost classifier model had the highest performance, with an AUC of 0.998 (95% CI: 0.995-1.000), and was therefore selected as the output model for this study. ConclusionsThe DL radiomics machine learning model based on MRI axial high-resolution T2WI images can effectively predict T2 and T3 stages of rectal cancer, with the XGBoost classifier model with combined features of the training set having the best performance. ]]> Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.015 ObjectiveTo explore the correlation and clinical value of arterial spin labeling (ASL) in quantitatively evaluating renal blood flow (RBF) and renal injury in patients with type 2 diabetes mellitus (T2DM). Materials and MethodsProspective recruitment of T2DM patients with varying degrees of renal function impairment was conducted. The patients were grouped based on the presence or absence of proteinuria and the estimated glomerular filtration rate (eGFR). Group A included T2DM patients without proteinuria, group B consisted of patients with proteinuria and eGFR≥60 mL/(min·1.73 m²), representing diabetic kidney disease (DKD) stages Ⅰ-Ⅱ, and group C comprised patients with 15 mL/(min·1.73 m²)≤eGFR＜60 mL/(min·1.73 m²), representing DKD stages Ⅲ-‍Ⅳ. Healthy volunteers were included as a control group. Dual-kidney routine MRI scans and ASL scans were performed to obtain RBF values in the cortical region of both kidneys. Additionally, biochemical indicators such as renal function and urine analysis were collected. Statistical analysis was conducted to compare the differences in cortical RBF values among the groups and assess the correlation, sensitivity, and optimal diagnostic thresholds of the relevant indicators. ResultsNo significant difference in cortical RBF values was found between the left and right kidneys (all P＞0.05). There was a significant overall difference in cortical RBF values among the four groups (all P＜0.001). Compared to the control group (174.28±23.89) mL/(100 g·min), group A exhibited a decrease to (159.66±28.54) mL/(100 g·min), group B decrease to (142.16±19.49) mL/(100 g·min), and group C showed the lowest value at (122.55±18.59) mL/(100 g·min). The reductions in RBF values for groups A, B, and C were 8%, 18%, and 30%, respectively. Group C had a significant decrease in cortical RBF values compared to groups A and B (both P＜0.05), with reductions of 23% and 14%, respectively. Group B also had a significant decrease compared to group A (P＜0.05), with a reduction of 11%. Cortical RBF values showed a negative correlation with serum creatinine (Scr) levels (r=-0.429, P＜0.001) and a positive correlation with eGFR (r=0.377, P＜0.001). The area under the curve (AUC) values of cortical RBF in distinguishing healthy volunteers from T2DM were 0.651 (95% CI: 0.520-0.768), distinguishing T2DM from DKD were 0.734 (95% CI: 0.619-0.829), and distinguishing early-stage DKD from mid-to-late-stage DKD was 0.760 (95% CI: 0.580-0.891). ConclusionsASL imaging provides a noninvasive quantitative evaluation of RBF changes and can effectively assess the progression of T2DM based on decreased RBF. It holds promise as an effective imaging method for noninvasively evaluating renal function damage in patients with T2DM. ]]> Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.016 ObjectiveTo study the feasibility of amide proton transfer (APT) imaging and apparent diffusion coefficient (ADC) in the preoperative histological grade assessment of bladder cancer. Materials and MethodsThe imaging data of 54 cases of bladder cancer confirmed by surgery and pathology were retrospectively analyzed, all patients underwent magnetic resonance examination before surgery, and the scanning sequences included APT and diffusion-weighted imaging (DWI) functional sequences except for conventional sequences. APT and ADC values of tumors were measured by two experienced radiologists independently. An independent samples t-test or Mann-Whitney U-test was used to compare the differences in each parameter between different grades of tumors. The receiver operating characteristic (ROC) curves and DeLong test were used to evaluate the efficiency of APT, ADC and combined parameters. ResultsThe APT values for low-grade bladder cancer (1.42%±0.75%) were significantly lower than high-grade bladder cancer (2.99%±1.07%), with statistically significant differences (P＜0.01); and the ADC values for low-grade bladder cancer were significantly higher [(1.44±0.27)×10^-3 mm²/s] than high-grade bladder cancer [(1.17±0.37)×10^-3 mm²/s], with statistically significant differences (P=0.01). The ROC curves showed that the area under the curve (AUC) of APT, ADC and the combined indicators for differentiating low-grade and high-grade bladder cancers were 0.80, 0.76, 0.94, and the AUC of combined parameters was significantly higher than that in APT or ADC (all P＜0.05). ConclusionsAPT and ADC can be used as indicators to predict the preoperative histological grades of bladder cancer, and the combination of APT and ADC can significantly improve the diagnostic efficacy. ]]> Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.017 ObjectiveTo investigate the feasibility of synthetic MRI [longitudinal relaxation time (T1), transverse relaxation time (T2), and proton density (PD)] for the prediction of the lymphovascular space invasion (LVSI) status in cervical squamous cell carcinoma (CSCC). Materials and methodsPatients who first went to the Second Hospital of Shanxi Medical University for suspected cervical cancer from May 2020 to November 2022 were prospectively collected. All patients underwent preoperative routine MRI scan, sagittal synthetic MRI to obtain the T1, T2, and PD values of the tumor. According to the LVSI status confirmed by postoperative pathological diagnosis, the subjects were divided into positive and negative LVSI groups. Use independent sample t-test or U-test to compare T1, T2 and PD values between the two groups. The receiver operating characteristic (ROC) curve was applied to evaluate the diagnostic efficacy of each parameter in predicting the LVSI status of CSCC. ResultsA total of 80 patients with operatively pathology confirmed CSCC were included in the study, including positive LVSI (n=51) and negative LVSI (n=29). There were significant differences in T1 value [(1191.60±101.17) ms vs. (1316.58±107.42) ms] and T2 value [(80.72±5.62) ms vs. (89.79±7.43) ms], all P＜0.001. In term of distinguishing LVSI positive from LVSI negative, the area under the curve (AUC) for T1 and T2 values were 0.798 and 0.850, respectively. A combination of T1 and T2 values showed a higher diagnostic performance (AUC=0.881), although there was no significant difference between T1, T2 and the AUC value of the combined parameter model (P＞0.05) by DeLong test. ConclusionThe quantitative parameters from synthetic MRI have potential for evaluating the LVSI status of CSCC and to help to optimize therapeutic strategies. ]]> Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.018 ObjectiveTo investigate the relationship between duration of illness and sternal stalk hypertrophy in patients with synovitis-acne-pustulosis-hyperostosis-osteitis (SAPHO) syndrome and to analyze their imaging features. Materials and MethodsImaging and clinical data of 24 patients with SAPHO syndrome diagnosed in our hospital from January 2021 to September 2022 were retrospectively analyzed. CT and MRI were performed on axial skeletal involvement, including the anterior thorax, spine, and sacroiliac joints. The patients^, disease duration was classified as ≤5 years and ＞5 years, divided into 14 cases and 10 cases. Measurements of the sternal stalk were made on chest CT images to explore the relationship between the duration of the disease and the hypertrophy of the sternal stalk. ResultsThe sites of involvement in 24 patients included the anterior chest wall, spine and sacroiliac joints. The main types of lesions included osteosclerosis, bone erosion, bone marrow edema, and fat deposition. The incidence of osteosclerosis and bone erosion lesions in the anterior chest wall was higher than that in the spine, and the difference was statistically significant (P=0.017, P=0.030), and the incidence of fat deposition lesions in the spine was higher than that in the anterior chest wall and sacroiliac joints, and the difference was statistically significant (P=0.001, P=0.017). Patients with a disease duration of ＞5 years had higher rates of sternal stalk thickness and sternoclavicular joint involvement than those with a disease duration of ≤5 years, and the difference was statistically significant (P=0.023, P=0.020). The changes in sternal stem width (r²=0.003, P=0.815) and sternal stem thickness (r²=0.035, P=0.379) were not correlated with the course of the disease. ConclusionsThe types of lesions in SAPHO syndrome are complex, with osteosclerotic and bony erosive lesions being more prevalent in the anterior chest wall and fat deposition lesions being more common in the spine. A long course of the disease is more likely to lead to a change of sternal stalk thickness, and the involvement of the sternoclavicular joint may be a sign of a long course of disease. ]]> Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.019 ObjectiveTo establish a joint nomogram model based on sacroiliac joint MRI radiomics features and clinical parameters to assist in the diagnosis of axial spondyloarthritis (axSpA). Materials and MethodsA retrospective analysis was performed for a cohort of 204 patients suspected of having axSpA, who visited our institution from April 2019 to September 2021. One hundred and two of these patients were diagnosed with axSpA and 102 were healthy controls. Their clinical features were subjected to univariate and multivariate analysis, and features with statistically significance were utilized to construct a clinical signature using machine learning algorithms. Regions of interest were delineated from the sacroiliac joint MRI T1-weighted (T1WI) and fat-suppressed T2-weighted (FS-T2WI) sequences taken by these patients, and radiomics features were extracted from these sequences. Intraclass correlation coefficient, Pearson correlation coefficient, least absolute shrinkage and selection operator were used to select features with strong relevance, and five machine learning models (logistic regression, RandomForest, ExtraTrees, XGBoost, multivariate logistic regression) were used to construct radiomics signatures for judging sacroiliac joint changes. Finally, the clinical and radiomics signatures were integrated to establish a comprehensive nomogram model. ResultsThe clinical signature was constructed using C-reactive protein and erythrocyte sedimentation rate. A total of 1834 radiomics features were extracted from each sequence of the sacroiliac joint MRI. After merging the features of different sequences, a total of 3368 features were obtained, from which the most relevant features were selected to construct radiomics signatures. For T1WI, FS-T2WI and fusion models, the best performing machine learning model was logistic regression. Radiomics signatures derived from fusion models displayed the best diagnostic performance. The final nomogram model exhibited excellent diagnostic performance in both the training set area under the curve (AUC) was 0.997 [95% confidence interval (CI): 0.992-1.000] and the testing set AUC was 0.944 (95% CI: 0.889-1.000). Decision curve also demonstrated that the nomogram model showed better predictive performance and clinical application value. ConclusionsThe nomogram model, incorporating sacroiliac joint MRI radiomics features and clinical parameters, demonstrates a strong capability to differentiate axSpA patients from healthy controls, which might facilitate clinical decision-making process. ]]> Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.020 ObjectiveTo investigate efficacy of radiomics on the lumbar spine MRI based on T2WI sequences in identifying osteoporosis. Materials and MethodsA retrospective analysis was conducted on a total of 291 patients who underwent lumbar spine MRI examinations at our hospital between December 2022 and March 2023. Regions of interest (ROI) were delineated layer by layer on the sagittal T2WI images. Radiomic features were extracted from the MR images of 1455 lumbar vertebrae. The samples were randomly divided into a training group (n=233) and a test group (n=58) at an 8∶2 ratio. The least absolute shrinkage and selection operator (LASSO) was used to reduce data dimensionality and select features. Logistic regression (LR) was employed to establish clinical models, radiomic models, and a combined model for predicting osteoporosis. The performance of the composite models was evaluated using metrics such as the area under the curve (AUC) of receiver operating characteristic (ROC), accuracy, specificity, sensitivity, positive predictive value, and negative predictive value. DeLong test was used to compare the predictive performance among the models. Calibration curves for the models were plotted, and Hosmer-Lemeshow test was applied to assess model fit. Decision curve analysis (DCA) was used to evaluate the clinical utility of each model. ResultsIn the training group, the AUCs for the clinical model, radiomic model, and combined model were 0.791 [95% confidence interval (CI): 0.733-0.849], 0.879 (95% CI: 0.833-0.925), and 0.893 (95% CI: 0.853-0.934), respectively. In the test group, the AUCs were 0.805 (95% CI: 0.676-0.935), 0.913 (95% CI: 0.841-0.985), and 0.904 (95% CI: 0.825-0.984), respectively. DeLong test results indicated that there was a statistically significant difference between the combined model and the clinical model (P＜0.05), while there was no statistically significant difference between the combined model and the radiomic model (P＞0.05). The Hosmer-Lemeshow test showed that the models were well calibrated (P=0.250, 0.753, 0.575). The results of DCA demonstrated that both the radiomic model and the combined model had better clinical value for predicting osteoporosis compared to the clinical model. ConclusionsAn image-based radiomics model constructed from lumbar T2WI has the potential for objective and accurate osteoporosis diagnosis. ]]> Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.021 ObjectiveTo investigate the value of MR intravoxel incoherent motion imaging (IVIM) in evaluating renal fibrosis (RF) in rabbits with renal artery stenosis. Materials and MethodsSeventy-eight healthy New Zealand white rabbits were randomly divided into control group (n=8) and RF group (n=70). The RF group was further randomly divided into six subgroups: pre-RF、RF-1W、RF-2W、RF-3W、RF-4W subgroup (n=12 in each subgroup), and RF-longitude (RF-L) subgroup (n=10) for dynamic observation. The rabbits in the RF group underwent left renal artery stenosis to establish the RF model. MR axial T2WI and IVIM scans were performed before operation and at 1, 2, 3 and 4 weeks after operation, respectively. After the last MR scan, the left kidney was resected and histopathological examination was performed. The rabbits in the RF-L subgroup and the control group underwent MR scanning at each time point and histopathological examinations at the end of the scanning at 4 weeks after operation. The true diffusion coefficient (D)、pseudo-diffusion coefficient (D^*) and perfusion fraction (f) values of bilateral renal cortex and medulla were measured and calculated on IVIM images. Repeated measures analysis of variance was used to compare the differences of D, D^* and f values in the left and right renal cortex and medulla over time in each subgroup. Independent sample t test was used to compare the differences of D, D^* and f values of bilateral renal cortex and medulla at each time point. Spearman correlation analysis was used to compare the correlation between the D, D^*, f values of the left renal cortex and medulla and the degree of RF. Receiver operating characteristic (ROC) curve was used to determine the efficacy of D, D^*, f values of left renal cortex and medulla in the diagnosis and staging of RF. ResultsIn the RF group, the D and f values of left renal cortex and medulla decreased gradually with RF. The D^* values of left renal cortex were significantly different between RF-4W and pre-RF, RF-1W, RF-2W subgroups (P all＜0.05). However, the D^* values of the left renal medulla did not differ significantly over time. The D^* and f values of the right renal cortex and medulla gradually increased with RF. However, the D values of the right kidney cortex and medulla did not differ significantly over time. In the control group, there was no significant difference in IVIM parameters between left and right renal cortex and medulla over time. The D and D^* values of the left renal cortex and medulla in the RF-4W subgroup were statistically different (P all ＜0.05). The f values of the left renal cortex and medulla in the pre-RF, RF-3W and RF-4W subgroups were statistically different (P all ＜0.05). The f values of the right renal cortex and medulla in the pre-RF and RF-4W subgroups were statistically different (P all ＜0.05). Spearman correlation test showed that the D and f values of left renal cortex and the f value of left renal medulla were moderately or strongly negatively correlated with the degree of RF (r=-0.595, -0.717, -0.412, P all ＜0.01). The D^* values of left renal cortex and the D、D^* values of the left renal medulla were not correlated with the degree of RF (P all ＞0.05). The f value was the best to identify pre-RF and RF-1W-RF-4W, pre-RF-RF-1W and RF-2W-RF-4W. The D value was the best to identify pre-RF-RF-2W and RF-3W-RF-4W, pre-RF-RF-3W and RF-4W. ConclusionsMR-IVIM can reflect the occurrence and development of RF from two aspects of perfusion and diffusion respectively. It has a great value in evaluating and staging RF. ]]> Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.022 ObjectiveTo investigate the feasibility of reducing the effect of Gd-diethylenetriamine pentametric acid (Gd-DTPA) dose on image quality in MRI brain enhancement by improving the acquired image through deep learning reconstruction (DL-Recon) algorithm. Materials and MethodsThe patients were divided equally into two groups, the normal dose group and the reduced dose group, and the corresponding T1WI were acquired using conventional reconstruction methods and DL-Recon techniques for the two groups. The region of interest was determined for signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) analysis under double-blind conditions by two associate chief, and subjective and objective evaluation of image quality, artefacts, homogeneity and enhancement effects were performed, with the subjective evaluation method based on the LIKERT guideline. The objective evaluation was performed by calculating the SNR of the image background, the superior frontal gyrus (SFG), the subarachnoid space (SAS) and the red nucleus (RN), respectively. The SNR was evaluated for image contrast. ResultsThe image quality obtained using DL-Recon was significantly increased (SNR_SFG increased by 48.9%, CNR_SFG increased by 91.5%) compared with that of conventional reconstruction method after injection of normal dose, and there was no significant correlation with the injection dose of Gd-DTPA. The artifacts and overall quality scores of DL-Recon were significantly higher than those of conventional images (P＜0.05). There was no significant difference in the enhancement effect of DL-Recon image + reduced dose group compared with that of conventional reconstruction in the normal injection dose group (P＞0.05). ConclusionsMRI DL-Recon algorithm has the ability to reduce the injection amount of Gd-DTPA on the premise of ensuring image quality. ]]> Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.023 ObjectiveTo investigate the value of diffusion weighted imaging (DWI), conventional MRI (cMRI) finding and demographic data to diagnose parotid gland tumors. Materials and MethodsA retrospective evaluation was made of the clinical data, histopathologic data, preoperative cMRI and DWI of 85 patients who underwent parotidectomy from November 2015 to September 2021 in Tongren People^,s Hospital. They were classified into three categories according to pathology: pleomorphic adenoma (PA), Warthin^,s tumor (WT) and malignant tumor (MT). Values of apparent diffusion coefficient (ADC) for PA, WT and MT were compared. The signal of lesion was observed on cMRI sequence. Classification variables were tested using Pearson^,s chi-squared test. For metric parameters, normality tests were performed first. If they follow a normal distribution, t-tests erre used, else non-parametric tests will be used. ResultsA total of 35 cases were PA, 31 cases were WT and the remaining 19 cases were MT. the PA group was younger and more frequent in females than the WT and MT groups, while WT and MT were more frequent in males. The PA group was younger and more frequent in females than the WT and MT groups, whereas it was more frequent in WT and MT in males. The cMRI was able to clearly visualize anatomical structures. The median of ADC values for them were (1.53±0.29) ×10^-3 mm²/s, (0.77±0.20) ×10^-3 mm²/s and (1.06±0.19) ×10^-3 mm²/s, respectively. PA was differentiated from the other two groups (P＜0.001). When the ADC threshold was 1.23×10^-3 mm²/s, the sensitivity and specificity for differentiating PA from WT was 94.3% and 93.5%, respectively; and when the ADC threshold was 1.29×10^-3 mm²/s for identifying PA from MT was 85.7% and 84.2%, respectively. ConclusionscMRI can provide anatomical and adjacent structural information for parotid gland tumors, and the ADC value can characterize the histopathological features of the tissue. The combination of the two is helpful for distinguishing between PA, WT, and MT, and is an important auxiliary tool for preoperative evaluation of parotid gland tumors. ]]> Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.025 Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.026 Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.027 Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.028 Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.029 Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.030 ^,s physical and mental health. Early diagnosis can significantly improve the survival rate of breast cancer patients. In recent years, with the development of big data and computer algorithms, the research and application of artificial intelligence (AI) such as radiomics and deep learning in the field of medical imaging have become increasingly extensive. It makes accurate and efficient imaging evaluation possible. The recent research on the status and progress of medical image-based AI in preoperative benign or malignant evaluation of breast cancer, breast cancer classification and histological grading, biomarkers and molecular subtyping prediction, pathological status of lymph nodes and susceptible gene diagnosis are reviewed in this article. The current status and problems of AI development in this field are reviewed and analyzed here to promote the clinical translation of AI technologies for breast cancer diagnosis and provide optimal radiological assistance for precise noninvasive clinical diagnosis and treatment. ]]> Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.031 Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.032 Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.033 Mon,20 Nov 2023 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2023.11.034 Mon,20 Nov 2023 00:00:00 GMT