http://med-sci.cn/cgzcx/en/contents_list.asp?issue=202601 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.1674-8034.2026.01.001 ObjectiveTo use three-dimensional arterial spin labeling (3D-ASL) imaging technology to measure cerebral blood flow (CBF) in very premature infants born to mothers with gestational diabetes mellitus (GDM) and to analyze its correlation with metabolic and inflammatory markers in these infants. Materials and MethodsOne hundred very premature infants were recruited from the Third Affiliated Hospital of Zhengzhou University between April 2024 and March 2025. The infants were divided into a GDM group consisting of 50 cases and a control group including 50 cases. Both groups performed 3D T1 BRAVO, 3D-ASL, and normal MRI sequence scans. General clinical data and CBF values across several brain regions were gathered and analyzed across the two groups. The correlation between CBF values in markedly distinct brain areas and the initial blood glucose levels and inflammatory markers was examined. ResultsIn comparison to the control group, the very premature infants in the GDM group had significantly higher birth weight, CBF values in multiple brain regions (left frontal lobe, right temporal lobe, bilateral occipital lobes, and bilateral basal ganglia), along with increased inflammatory markers (P < 0.05). The initial blood glucose levels in the GDM group were markedly lower than those in the control group (P < 0.05). The CBF values in the brain regions exhibiting differences between the two groups were positively correlated with procalcitonin levels (r = 0.665, 0.518, 0.627, 0.582, 0.495, 0.465, P < 0.05) and negatively correlated with initial blood glucose levels (r = -0.409, -0.448, -0.450, -0.447, -0.487, -0.501, P < 0.05). Conclusions3D-ASL technology offers a noninvasive and dependable imaging modality for assessing brain development in very premature infants born to mothers with GDM. Alterations in blood flow parameters connect with clinical signs, providing significant references for the early detection and personalized management in these newborns, perhaps enhancing their prognosis. ]]> Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.002 ObjectiveTo detect the alterations of spontaneous brain activity and functional connectivity (FC) in chronic smokers employed the method of amplitude of low-frequency fluctuation (ALFF) combined with FC. Materials and MethodsA total of 27 chronic smokers and 30 healthy controls were enrolled in this current study, which was matched with age and gender. All participants underwent neuropsychological assessments and resting-state functional magnetic resonance imaging (rs-fMRI) scans. The DPABI 6.0 and SPM 12.0 software were used to analyze the brain regions with abnormal ALFF values. The brain region with abnormal ALFF values was selected as the see region of interest (ROI) for the whole brain functional connectivity analysis. ResultsCompared with the health controls, the chronic smokers exhibited increased ALFF value in the left superior frontal gyrus (SFG) and a decreased ALFF value in the right cerebellar hemispheres. Additionally, the current study found increased FC between the left SFG and the left parietal gyrus, and we also found increased FC between the left SFG and the right parietal gyrus (GRF correction, voxel-level P < 0.001 cluster-level P < 0.05). ConclusionsAbnormal brain activity and functional connectivity of the frontal gurus, parietal gyrus and cerebellum may be the reason of addiction in chronic smokers, which may provide imaging evidence to reveal the neurobiological mechanisms in chronic smokers. ]]> Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.003 ObjectiveTo explore the variation characteristics of gray matter volume (GMV) and GMV-based structural covariance network (SCN) of subcortical nuclear structures in patients with classic trigeminal neuralgia (CTN). Materials and MethodsThe 3D-T1WI structural image data of 55 patients with classic CTN and 59 healthy controls were prospectively collected, and the GMV of bilateral thalamus, hippocampus and amygdala subregions was extracted, and calculate the SCN at the group level based on the GMV of each subject. Compare the differences in GMV and SCN parameters between the two groups respectively. ResultsCTN patients showed volume reduction in the left anteriorventral, paracentral, parafascicular, right laterodorsal, central medial, reuniens medial ventral and bilateral ventromedial in the subregions of the thalamus (P < 0.001). In the amygdala subregion, the volumes of the right accessory basal nucleus, the anterior amygdaloid area, the cortico-amygdaloid transition area, and the bilateral cortical nucleus (P < 0.001) decreased; in the hippocampal subregion, the volume of the left cornu ammonis 4-body and the granule cell layer of the dentate gyrus-body (P ≤ 0.001) decreased. Partial correlation analysis indicated that the duration of pain in the CTN patient group was negatively correlated with the volumes of the left ventromedial, the left parafascicular, the right ventromedial and the right central medial (r_s = -0.397, P = 0.003; r_s = -0.435, P < 0.001; r_s = -0.306, P = 0.023 and r_s = -0.323, P = 0.016, respectively), the score of the self-rating anxiety scale was positively correlated with the volume of the right cortico-amygdaloid transition area and the right accessory basal nucleus (r_s = 0.257, P = 0.059 and r_s = 0.349, P = 0.009, respectively). The small-world index of SCN in the CTN group was significantly lower than that in the control group (P < 0.05). ConclusionsThe important subregions of subcortical nuclear structure in CTN patients show volume reduction and changes in network properties. These findings suggest that the alterations in subcortical nuclear structure and structural covariance network characteristics, as potential structural feature markers of CTN, are expected to provide new targets for pain treatment. ]]> Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.004 ObjectiveTo investigate brain functional network changes in adolescents with depression and childhood trauma after eye movement desensitization and reprocessing (EMDR) combined with sertraline versus sertraline monotherapy, and evaluate the advantages of combined therapy. Materials and MethodsA total of 67 adolescent patients with depression and childhood trauma were enrolled. General demographic data and Hamilton Depression Rating Scale (HAMD) scores were collected. Participants were divided into two groups: 33 receiving EMDR combined with sertraline and 34 receiving sertraline monotherapy. All underwent 8-week treatment. Clinical evaluations included the Hamilton Depression Rating Scale (HAMD). Pre- and post-treatment resting-state and structural imaging data were acquired. Functional connectivity matrices were constructed using average blood oxygen level-dependent (BOLD) signals from 142 regions of interest (ROI) defined by the cerebellum-removed Dosenbach atlas. Paired-sample t-tests compared intergroup differences in functional connectivity matrices. Differential edges were categorized into large-scale brain networks using the Yeo7 model, and the number of differential edges within/between seven networks was calculated. ResultsBoth groups showed significant HAMD score reductions (combined therapy: t = 16.11, P < 0.001; monotherapy: t = 13.20, P < 0.001) and widespread brain functional connectivity changes. The combined therapy group had 49 differential ROIs and 39 differential edges (all P < 0.05), with reduced connectivity primarily within the default mode network (DMN) and ventral attention network (VAN), and increased connectivity between DMN-frontoparietal network (FPN) and DMN-dorsal attention network (DAN). The sertraline group had 32 differential ROIs and 19 differential edges (all P < 0.05), with reduced connectivity mainly within DMN and increased connectivity involving DAN. ConclusionsBoth treatments demonstrated effective antidepressant outcomes. Combined therapy induced enhanced DMN-FPN connectivity, suggesting improved coordination between cognitive flexibility and emotion regulation, with broader therapeutic effects compared to monotherapy. ]]> Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.005 ObjectiveThrough voxel-based morphometry (VBM), the structural magnetic resonance imaging (sMRI) features of schizophrenia (SCZ) patients with and without aggressive violence were compared, and a machine learning model was constructed to realize the early identification and prediction of SCZ patients with aggressive violence. Materials and MethodsA retrospective analysis of 146 patients diagnosed with SCZ in Shandong Daizhuang Hospital from March 2023 to June 2025, including 77 SCZ patients with aggressive violence and 69 SCZ patients without aggressive violence. The differences of clinical indicators and sMRI features between the aggressive violence group and the non-aggressive violence group were compared, and the prediction model of aggressive violence in SCZ patients was constructed. Among the four machine learning prediction models, the area under the curve (AUC) of logistic regression (LR), decision tree (DT), random forest (RF) and support vector machine (SVM) were 0.824, 0.821, 0.917 and 0.940, respectively. The results of DeLong test showed that the AUC of LR, DT, RF and SVM were 0.824, 0.821, 0.917 and 0.940, respectively. The predictive performance of the SVM model was the highest (P < 0.05). The SHAP (SHapley Additive exPlanations) summary map results showed that the thickness of the left insular cortex was the most important feature for predicting aggressive violence in schizophrenia. Decision curve analysis (DCA) showed that the four models had high guiding significance for clinical practice. ResultsThere was no significant difference in clinical indicators between the aggressive violence group and the non-aggressive violence group (P > 0.05). The thickness of the left insular cortex, the thickness of the left fusiform cortex, the Gaussian curvature of the left operculum, the thickness of the right insular cortex, the mean curvature of the left operculum, and the thickness of the right middle and posterior prefrontal cortex in the aggressive violence group were lower than those in the non-aggressive violence group (P < 0.05). ConclusionsMachine learning models based on sMRI can predict aggressive and violent behaviors in SCZ patients, with the SVM model exhibiting the highest predictive performance. ]]> Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.006 ObjectiveTo investigate the relationship between arterial recanalization after endovascular treatment and the length and width of the susceptibility vessel sign (SVS), and to evaluate the relationship between successful arterial recanalization after endovascular treatment and various risk factors. Materials and MethodsWe retrospectively analyzed the patients with anterior circulation ischemic stroke who underwent magnetic resonance imaging (MRI) of the head and susceptibility-weighted imaging (SWI) before endovascular treatment, and measured the SVS length and width. To reduce the variation of SVS width caused by different parts and individual differences, divide the width of the SVS on the affected side by the width of the healthy blood vessel to obtain the relative width of the SVS. At the end of endovascular treatment, those who achieved an assessment of 2b50-3 in the expanded thrombolysis in cerebral infarction (eTICI) were classified as the successful recanalization group, while those who did not reach this level were classified as the unsuccessful recanalization group. Stepwise regression was used to screen variables, and the indicators with statistically significant differences were included in the multivariate logistic regression analysis to determine the independent predictors of successful recanalization and first-pass reperfusion (FPR) after endovascular treatment. ResultsAmong 112 patients, 79 cases achieved successful recanalization and 33 cases did not,within the successful recanalization group, there were 39 cases of FPR. The median width of SVS in the successful recanalization group and the unsuccessful recanalization group was 4.3 (4.1, 4.6) mm and 3.2 (3.0, 3.9) mm, respectively, with P < 0.001, indicating a statistically significant difference; the median length of SVS was 10.6 (9.5, 13.1) mm and 10.5 (7.5, 14.5) mm, respectively, with P = 0.871, indicating no statistically significant difference. In the multivariate logistic regression analysis, SVS width was associated with successful recanalization [odds ratio = 3.025; 95% confidence interval (CI): 4.895 to 24.564; P = 0.001] and FPR (odds ratio = 9.243; 95% CI: 3.493 to 14.460; P < 0.001). The optimal cutoff value for the SVS width to predict successful recanalization and FPR were 3.95mm and 3.85 mm, respectively. ConclusionsThe width of SVS is a potential imaging biomarker for predicting successful reperfusion of arteries and FPR after endovascular treatment. The length of SVS has no significant predictive value. ]]> Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.007 ObjectiveTo explore the value of dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI)-based tumor heterogeneity quantification integrated with deep learning (DL) in predicting the pathological complete response of neoadjuvant chemotherapy (NAC) for breast cancer. Materials and MethodsThe clinical and imaging data of 179 patients with pathologically confirmed breast cancer at the First Affiliated Hospital of Wannan Medical College from January 2019 to January 2025 were retrospectively collected. Among them, 58 patients achieved pathological complete response (pCR) after NAC, and 121 patients achieved non-pathological complete response (non-pCR). The patients were randomly divided into a training group (n = 125) and a validation group (n = 54) at a ratio of 7∶3. All patients underwent MRI examination before NAC. The ITK-SNAP software was used to manually delineate the region of interest (ROI) layer by layer and perform three-dimensional fusion. The Gaussian mixture model (GMM) was used for cluster analysis, and the Bayesian information criterion (BIC) was used to determine the sub-regions of the tumor lesions. The intratumoral heterogeneity score (ITH-score) was calculated, and a habitat imaging model was established. The PyRadiomics package in Python software was used to extract the traditional radiomics features of the whole tumor, and the ViT deep learning model was used to extract the deep learning features. The minimum redundancy maximum relevance (mRMR) and least absolute shrinkage and selection operator (LASSO) regression methods were used for feature dimensionality reduction and screening. A traditional radiomics model and a deep learning model were constructed respectively, and the quantitative score of each patient was calculated according to the feature weights in the models. Multivariate logistic regression analysis was used to construct a clinical model and a combined model. Receiver operating characteristic (ROC) curves were drawn to evaluate the predictive efficacy of each model. The DeLong test was used to compare the efficacy of each model, and decision curve analysis (DCA) was used to analyze the clinical benefits of the models. The SHAP method was used to analyze the importance of each feature in the combined model. ResultsThe AUC [95% (confidence interval, CI)] values of the clinical model, traditional radiomics model, deep learning model, habitat imaging model, and combined model in predicting pCR after NAC in the training group were 0.864 (0.832 to 0.895), 0.776 (0.745 to 0.807), 0.728 (0.703 to 0.752), 0.823 (0.785 to 0.881), and 0.943 (0.903 to 0.983) respectively, and in the validation group were 0.732 (0.684 to 0.781), 0.634 (0.589 to 0.679), 0.757 (0.720 to 0.791), 0.750 (0.690 to 0.840), and 0.875 (0.821 to 0.929) respectively. The combined model had the best predictive performance. The DCA results showed that the clinical benefit of the combined model was higher than that of the clinical model and other radiomics models. In the SHAP method, the importance of the ITH-score was higher than that of the molecular subtype. The larger the SHAP value, the more the prediction result tended to pCR. ConclusionsThe combined model based on DCE-MRI heterogeneity quantitative analysis and deep learning demonstrates superior predictive performance for pCR in breast cancer patients after NAC, which holds clinical application value for early prediction of pCR after NAC and contributes to clinical diagnosis and treatment management of breast cancer. ]]> Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.008 ObjectiveTo explore the diagnostic value of gadoxetic acid disodium (Gd-EOB-DTPA) enhanced MRI in differentiating of focal chronic inflammatory hepatic lesions (FCIHL) from solitary hepatic metastasis (SHM). Materials and MethodsThe imaging and clinical data of 23 patients with FCIHL and 36 patients with SHM pathologically confirmed in the Shunde Hospital of Guangzhou University of Chinese Medicine from June 2019 to February 2024 were retrospectively collected. All patients underwent diffusion-weighted imaging (DWI) and gadoxetic acid disodium (Gd-EOB-DTPA) enhanced MRI examinations. The imaging characteristics and clinical indicators were compared between the two groups. Kappa test or intra-class correlation coefficient (ICC) was used to assess the consistency between two observers in each image interpretation. For clinical and imaging data of the two groups that conformed to normal distribution, independent samples t-test was adopted; for other continuous data, Mann-Whitney U test was applied; and for categorical data, chi-square test was used to analyze inter-group differences. Logistic regression analysis was conducted on variables with significant inter-group differences (P < 0.05) to identify the independent risk factors for FCIHL. A nomogram was constructed, and receiver operating characteristic (ROC) curves were plotted to obtain the area under the curve (AUC) values. ResultsInterobserver consistency between the two observers was excellent, with all ICC or Kappa values greater than 0.8 (all P < 0.05). The apparent diffusion coefficient (ADC) value and the diameter difference of lesions between T2WI and hepatobiliary phase were independent influencing factors for differentiating FCIHL from SHM, with AUC values of 0.787 (95% CI: 0.671 to 0.904) and 0.836 (95% CI: 0.737 to 0.936), respectively (both P < 0.05). The comprehensive AUC value of the nomogram is higher than that of a single indicator, reaching 0.908 (95% CI: 0.836 to 0.980). ConclusionsGd-EOB-DTPA-enhanced MRI has certain value in differentiating FCIHL from SHM. Among these imaging features, the diameter difference of lesions between T2WI and hepatobiliary phase is the most valuable discriminative indicator. ]]> Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.009 ObjectiveTo establish a prediction model based on multiparametric MRI radiomics and clinical-radiology features, and evaluate its efficacy in predicting neoadjuvant chemoradiotherapy (nCRT) in patients with locally advanced rectal cancer. The Shapley algorithm was employed to enhance model interpretability. Materials and MethodsA retrospective analysis was conducted on 172 patients who received nCRT and surgery from the First Affiliated Hospital of the University of Science and Technology of China (center 1) and the Hefei Cancer Hospital of the Chinese Academy of Sciences (center 2), and Clinical and MRI data were analyzed. According to the 8th edition AJCC tumor regression grading (TRG) criteria for rectal cancer, patients with TRG 0-1 were classified as good responders (GR), while those with TRG 2-3 were classified as poor responders (PR) based on postoperative pathological results. The GR group comprised 77 patients, and the PR group comprised 95 patients. Patients from center 1 were randomly divided into a training set (n = 92) and an internal validation set (n = 40), while the patients from center 2 were utilized as an independent external validation set (n = 40). High-resolution axial T2WI, diffusion-weighted imaging (DWI) and sagittal contrast-enhanced T1WI (CE-T1WI) sequences were selected to delineate the region of interest (ROI) along the tumor margins. PyRadiomics software was used to extract all radiomics features after image preprocessing. Spearman correlation analysis and least absolute shrinkage and selection operator (LASSO) analysis were used to retain the radiomics features strongly associated with the efficacy of nCRT. T2WI, DWI, CE-T1WI and multiparametric radiomics score (Rad-score) were obtained by eXtreme gradient boosting (XGBoost) classifier. The independent clinical-radiology predictors were screened by single-multiple logistic regression to build the clinical-radiology model, and the multiparametric model Rad-score combined with independent clinical-radiology predictors was selected to build the nomogram model. The performance of the model was evaluated using receiver operating characteristic (ROC) curves. The best-performing radiomics model was explained by the Shapley algorithm. ResultsUnivariate and multivariate logistic regression analysis identified age, tumor longest diameter, and neoadjuvant treatment modalities as independent predictors for treatment efficacy. The clinical-radiology model demonstrated the area under the curve (AUC) of 0.80 (95% CI: 0.75 to 0.85) in the training set, 0.73 (95% CI: 0.68 to 0.78) in the internal validation set, and 0.60 (95% CI: 0.55 to 0.65) in the external validation set. Among radiomics models, the multiparametric radiomics model (T2WI + DWI + CE-T1WI) achieved optimal performance, with AUCs of 0.98 (95% CI: 0.95 to 1.00), 0.95 (95% CI: 0.91 to 0.99), 0.86 (95% CI: 0.81 to 0.91) in the training, internal validation, and external validation sets, respectively. The nomogram model achieved the best predictive performance. The AUC, accuracy, sensitivity, and specificity of the training set of nomogram model were 0.99 (95% CI: 0.97 to 1.00), 98%, 95%, and 98%, respectively. The internal validation sets were 0.98 (95% CI: 0.95 to 1.00), 98%, 98% and 98%, respectively. The external validation sets were 0.88 (95% CI: 0.83 to 0.93), 88%, 87% and 87% respectively. DeLong test indicated that the nomogram model^,s performance was superior to the clinical model and the radiomics models (P < 0.05). Shapley analysis revealed that wavelet-LHL_glszm_SmallAreaEmphasis in DWI sequence was the most important feature in the radiomics model. ConclusionsThe nomogram based on multiparametric MRI radiomics and clinical-radiology features may be used as an accurate and non-invasive method to predict the efficacy of nCRT in rectal cancer patients, and the Shapley algorithm can provide interpretability of radiomics model. This nomogram has been validated using an external validation set, suggesting its potential utility of providing important guidance for clinical diagnosis and treatment decision-making. ]]> Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.010 ObjectiveTo construct an interpretable integrated model based on dynamic contrast-enhanced MRI (DCE-MRI) habitat imaging radiomics and clinical features, and to assess its utility in predicting lymph node metastasis (LNM) status in rectal cancer. Materials and MethodsA retrospective analysis was conducted on the clinicopathological and imaging data of 148 patients with rectal cancer admitted to Gansu Provincial People^,s Hospital between January 2016 and July 2024. Patients were stratified into LNM-positive and LNM-negative groups based on postoperative pathological confirmation. They were then randomly divided into a training cohort (n = 103) and a test cohort (n = 45) in a 7∶3 ratio. The region of interest (ROI) was manually delineated on the DCE-MRI parametric map using ITK-SNAP software. Subsequently, 19 standardized radiomics features were extracted from the K^trans maps. K-means clustering (K = 4) was applied to partition the tumor into distinct habitat subregions. Radiomics features were extracted separately from each tumor subregion (habitat-specific features) and from the whole tumor volume (whole-tumor features). The intra-class correlation coefficient (ICC) was calculated to assess the reproducibility of the whole-tumor radiomics feature extraction. Feature selection involved Z-score normalization, correlation analysis, and the least absolute shrinkage and selection operator (LASSO) algorithm. Predictive models for LNM status were developed using four machine learning classifiers: extremely randomized trees, logistic regression, random forest, and support vector machine. These models were built based on habitat-specific radiomics features and whole-tumor radiomics features separately. Logistic regression was also used to identify independent clinical predictors and construct a clinical model. Finally, an integrated model was built by combining significant clinical predictors with the radiomics signature derived from the habitat analysis. Model performance was evaluated using the receiver operating characteristic (ROC) curve and quantified by the area under the curve (AUC). Decision curve analysis (DCA) was performed to assess the clinical utility of the models. The importance of features in the final integrated model was determined, and the model^,s predictions were explained visually using Shapley additive explanations (SHAP) analysis. ResultUnivariate analysis identified carcinoembryonic antigen (CEA) level and MRI-reported N-stage as significant predictors of lymph node status in the training cohort [odds ratios (OR) = 2.346 and 7.727, respectively; 95% confidence intervals (CI): 1.052 to 5.233 and 2.273 to 26.268, respectively; P < 0.05]. The predictive model based on habitat radiomics features demonstrated superior performance, with AUC values of 0.890 (training cohort) and 0.801 (test cohort), outperforming the whole-tumor radiomics model (AUC: 0.774 training, 0.684 test). The integrated model, combining clinical features with the habitat radiomics signature, achieved the highest AUC values: 0.896 in the training cohort and 0.866 in the test cohort. DCA indicated that the integrated model provided a higher net clinical benefit across a range of threshold probabilities. SHAP analysis provided quantitative interpretability for the integrated model^,s predictions, revealing the habitat radiomics score as the most significant predictor. ConclusionsThe interpretable integrated model, constructed using preoperative DCE-MRI habitat imaging radiomics features and clinical factors, accurately predicts lymph node status in rectal cancer patients. By providing visual interpretation of individual predictions through SHAP, this model offers a valuable tool to support personalized treatment decision-making. ]]> Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.011 ObjectiveTo evaluate the predictive efficacy of virtual magnetic resonance elastography (vMRE) based on diffusion-weighted imaging (DWI) for lymph node metastasis (LNM) in cervical cancer patients undergoing direct surgery. Materials and MethodsClinical and imaging data of cervical cancer patients who underwent preoperative pelvic MRI and radical hysterectomy at Henan Provincial People^,s Hospital between November 2021 and November 2022 were retrospectively collected and analyzed. The pelvic MRI protocol included multi-b-value DWI, and vMRE images were generated from DWI data to extract the (μ_Diff) parameter. Based on postoperative pathology, patients were divided into LNM-positive and LNM-negative groups. The t-test or Mann-Whitney U test was used to compare differences in DWI-based virtual shear modulus μ_Diff parameters between groups, and logistic regression analysis was performed to identify variables associated with lymph node status. Predictive models were constructed, and receiver operating characteristic (ROC) curves were plotted. The area under the curve (AUC) was used to evaluate the predictive performance of each model. ResultsAmong clinical variables, squamous cell carcinoma antigen (SCCAG) and maximum lymph node short-axis diameter were significantly associated with LNM. The mean, maximum, and median μ_Diff values in the LNM-positive group were significantly higher than those in the negative group (P < 0.05). The combined model incorporating the maximum μ_Diff value and maximum lymph node short-axis diameter demonstrated the best predictive performance for LNM, with an AUC of 0.824 (95% CI: 0.683 to 0.965), superior to the single model constructed solely based on the mean μ_Diff value and the short-axis diameter of the largest lymph node. ConclusionsvMRE image features based on multi-b-value DWI can serve as a noninvasive indicator reflecting tissue stiffness, improving the predictive accuracy of LNM in cervical cancer patients. This approach provides a novel imaging biomarker for the preoperative noninvasive assessment of LNM. ]]> Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.012 ObjectiveTo establish a combined model based on MRI-based radiomics features and clinicopathological characteristics for evaluating the programmed death-ligand 1 (PD-L1) level in cervical cancer. Materials and MethodsA retrospective analysis was conducted on 327 cervical cancer patients who underwent MR enhanced scans at Liaoning Cancer Hospital & Institute, from January 2021 to September 2024. The samples were randomly divided into a training set (n = 228) and a validation set (n = 99) in a 7∶3 ratio. The PD-L1 combined positive score (CPS) ≥ 10 was used as the cut-off value and divided the patients into high and low expression groups. Radiomics feature selection was generated through the χ² test, the analysis of variance and random forest. An extreme gradient boosting (XGBoost) classifier was employed for model construction. Univariate logistic regression analysis was used to analyze the clinicopathological data. Radiomics modesl, clinicopathological models and combined models were developed for predicting the level of PD-L1. The predictive performance of the model was evaluated using the receiver operating characteristic (ROC) curve, calibration curve and decision curve analysis (DCA). ResultsThere were significant differences in human papillomavirus (HPV) infection and degree of differentiation between the high and low PD-L1 cervical cancer expression groups (all P < 0.05). The AUC of the clinical model in the training and validation sets were 0.672 [95% confidence interval (CI): 0.598 to 0.745] and 0.698 (95% CI: 0.578 to 0.819), respectively. Seven radiomics features were selected from 2261 extracted radiomics features to construct the model, and the AUC was 0.788 (95% CI: 0.728 to 0.848) and 0.712 (95% CI: 0.593 to 0.832) in the training and validation sets, respectively. The AUC of the combined model in the training and validation sets were 0.932 (95% CI: 0.898 to 0.967) and 0.805 (95% CI: 0.694 to 0.916), respectively. ConclusionsPD-L1 expression can be effectively predicted using an MRI-based radiomics model combined with clinicopathological characteristics to identify patients who may benefit from anti-PD-L1 immunotherapy. ]]> Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.013 ObjectiveTo construct a machine learning model for predicting the prognosis of concurrent chemoradiotherapy (CCRT) in patients with locally advanced cervical cancer (LACC) based on magnetic resonance imaging (MRI) radiomics features, and to evaluate its predictive performance. Materials and MethodsA retrospective analysis was performed on 424 LACC patients admitted to the Affiliated Cancer Hospital of Xinjiang Medical University from February 2019 to February 2020. Patients were randomly assigned to a modeling group (n = 339) and an internal validation group (n = 85) at a 4∶1 ratio. Additionally, 120 LACC patients admitted to the Second Affiliated Hospital of Xinjiang Medical University during the same period were enrolled as the external validation group. Clinical data and MRI images (including transverse dynamic contrast-enhanced T1WI, transverse fast spin echo T2WI, and transverse diffusion-weighted imaging sequences) were collected. The region of interest (ROI) was delineated in the lesion area, and radiomics features were extracted using PyRadiomics. Dimensionality reduction and selection of radiomics features were conducted via the least absolute shrinkage and selection operator (LASSO) algorithm. A radiomics model was constructed based on the selected features, and radiomics scores (Rads) were calculated. Cox univariate and multivariate analyses were performed using patient clinical data and Rads to establish a prognostic prediction model for CCRT in LACC patients, followed by validation of the model^,s predictive performance. ResultsThirteen MRI radiomics features were selected by the LASSO algorithm. Univariate and multivariate Cox analyses demonstrated that external beam radiotherapy dose [hazard ratio (HR) = 1.275, 95% CI: 1.186 to 1.371] and 2 Gy fractionated radiation equivalent dose (HR = 1.092, 95% CI: 1.050 to 1.137) were independent risk factors for mortality following CCRT in LACC patients, whereas hemoglobin (HR = 0.962, 95% CI: 0.940 to 0.985) and Rads (HR = 0.949, 95% CI: 0.933 to 0.965) were protective factors (all P < 0.05). Model validation showed that the area under the curve (AUC) values for internal and external validation were 0.978 (95% CI: 0.965 to 1.000) and 0.971 (95% CI: 0.958 to 0.996), respectively. The Hosmer-Lemeshow test yielded chi-square values of 8.580 (P = 0.379) and 8.691 (P = 0.370) for internal and external validation, respectively. ConclusionsThis study established a nomogram prediction model based on MRI radiomics, which exhibits excellent predictive performance and clinical utility for the prognosis of CCRT in LACC patients. It may serve as a reference for the formulation and adjustment of CCRT treatment plans for LACC patients. ]]> Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.014 ObjectiveTo establish a prenatal MRI-clinical scoring system and to explore its predictive value for the classifications of placenta accreta spectrum (PAS). Materials and MethodsWe retrospectively collected the clinical and imaging data of pregnant women who visited the obstetrics clinic of our hospital and were suspected of PAS based on ultrasound or clinical screening, from January 2018 to June 2023. PAS disorders were diagnosed and classified by the surgical and pathological examinations. According to the updated joint consensus of the American Society of Abdominal Radiology and the European Society of Urogenital Radiology, 11 MRI signs related to PAS were selected. The clinical independent risk factors for PAS were selected by univariate and multivariate logistic regression. The logistic regression analysis combining independent risk factors and all MRI signs was used to screen the features with P < 0.05. The weighting of each feature was calculated based on its β coefficient. The receiver operating characteristic (ROC) curve was used to evaluate the predictive performance of the scoring system, and the threshold of the score for each classification, area under the curve (AUC), sensitivity, specificity, and other diagnostic indicators were calculated. ResultsA total of 404 pregnant women who met the inclusion and exclusion criteria were collected. Seven features based on MRI and clinics were included in our scoring system. For the comparison between the non-PAS and PA groups, the cut-off value was 6.0. The optimal cut-off value between PA and PI groups was 11.0. The threshold value for PI and PP groups was 17.0. ConclusionsThe prenatal MRI signs-clinical scoring system possesses favorable clinical feasibility. It plays an important role for assisting in the diagnosis of PAS disorders and identifying the high-risk patients. ]]> Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.015 ObjectiveTo investigate the radiomic features of the lateral pterygoid muscle, articular disc, and bilaminar zone (retrodiscal tissue) on magnetic resonance imaging (MRI) in patients with temporomandibular disorder (TMD). The study aims to facilitate the early diagnosis and differential diagnosis of TMD subtypes, thereby enhancing clinical efficiency. Materials and MethodsThis retrospective study included imaging and clinical data from patients clinically diagnosed with TMD at the First Affiliated Hospital of University of Science and Technology of China between December 2019 and October 2024. A total of 121 patients were enrolled. Data from each joint side were analyzed independently, resulting in three groups: 33 sides with reducible disc displacement (RDD), 89 sides with non-reducible disc displacement (NRDD), and 120 sides without disc displacement (ND). Radiomic features were extracted from the lateral pterygoid muscle, articular disc, and bilaminar zone. Univariate analysis of variance (ANOVA) with Bonferroni correction was employed to identify features with significant differences among the three groups. Subsequently, a "One-vs-Rest" (OVR) strategy was used to train logistic regression models, with the selected significant features as inputs and the predicted probabilities for each group as outputs. Three predictive models were developed: a clinical model, a radiomics model, and a combined model (integrating both clinical and radiomic features), using univariate and multivariate logistic regression analyses. The training set was partitioned using stratified random sampling, and the diagnostic performance was evaluated on the testing set. After performing stratified 5-fold cross-validation, the model performance was evaluated by the area under the curve (AUC) of the receiver operating characteristic (ROC) curve. The DeLong test was used to assess the statistical significance of differences between AUCs, and decision curve analysis (DCA) was conducted to evaluate the predictive net benefit of the models. Results(1) Seven radiomic features from the bilaminar zone and five from the articular disc were identified as significantly different among the three groups. No significant features were found in the lateral pterygoid muscle, suggesting limited radiomic differences in this muscle across TMD subtypes. (2) The ROC curve analysis demonstrated that the significant features from the articular disc achieved high diagnostic performance in distinguishing between NRDD (AUC = 0.79) and ND (AUC = 0.78). The significant features from the bilaminar zone showed high efficacy in identifying ND (AUC = 0.84). (3) All statistical tests were two-sided, with the significance level set at 0.05. Evaluation on the testing set revealed that the combined model (AUC = 0.90) exhibited significantly stronger and more stable discriminatory power, as well as superior predictive and diagnostic efficacy, compared to the clinical model (AUC = 0.88) and the radiomics model (AUC = 0.78). ConclusionsWhile radiomic analysis of the lateral pterygoid muscle alone is insufficient for TMD subtype differentiation, our study establishes that MRI-based radiomics of the articular disc and bilaminar zone provides a novel paradigm for early diagnosis, differential diagnosis, and guiding clinical interventions in TMD. ]]> Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.016 ObjectiveTo investigate changes in signal intensity (SI) of the dentate nucleus (DN) on unenhanced T1-weighted magnetic resonance imaging (MRI) scans after multiple administrations of linear gadolinium-based contrast agents (GBCAs), and to analyze its correlation with various clinical factors. Meterials and Methods: Clinical and imaging data of ninety-two patients who underwent at least three consecutive linear GBCA-enhanced MRI examinations at our hospital from January 2015 to December 2024 were analyzed retrospectively. Unenhanced MRI scans were performed before and after consecutive enhanced MRI examinations in all patients. On a post-processing workstation, the mean SI of the DN and the pons were measured on unenhanced T1-weighted images. The SI ration of DN-to-pons was calculated by dividing the SI in the DN by that in the pons. A generalized additive model (GAM) was used to examine the trends and patterns of the SI ratio of DN-to-pons relative to the number of GBCAs administrations. Linear regression analysis was used to examine SI ratio of DN-to-pons correlation with various clinical factors. The incremental changes in the SI ratio of DN-to-pons between consecutive examinations were compared to analyze their trend and a trend analysis was used on the variation pattern. ResultsThe SI ratio of DN-to-pons increased with the cumulative number of linear GBCAs injections, following a non-linear pattern. The SI ratio differences showed a significant correlation with the number of injections (P < 0.001). There was no correlation with other clinical factors (P > 0.05). Analysis of the incremental changes in the ratio after the first 6 enhancements revealed median increments of 1.91, 0.94, 0.93, 0.88, 0.91, and 0.87. A trend test was performed on the incremental changes, revealing statistically significant differences (coefficient of the linear mixed-effects model: -0.215, P < 0.001; coefficient of the polynomial trend analysis: -7.530, P < 0.001). ConclusionsSerial injections of linear GBCAs may lead to a non-linear increase in SI in the DN,which correlates with the number of contrast-enhanced examinations, while the rate of increase gradually slows down as the number of such examinations rises. ]]> Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.019 Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.020 Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.021 Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.022 Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.023 Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.024 Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.025 Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.026 Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.027 Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.028 Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.029 Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.030 Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.031 Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.032 Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.033 Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.034 Tue,20 Jan 2026 00:00:00 GMT http://med-sci.cn/cgzcx/en/en_articlexml.asp?doi=10.12015/issn.1674-8034.2026.01.035 Tue,20 Jan 2026 00:00:00 GMT