Objective The aim of this study was to investigate frequency-specific alterations in spontaneous brain activity during the resting state in patients with AUD and to assess the correlation between these alterations and clinical data. Method Thirty-three patients with AUD and 29 healthy controls (HC) matched for age, sex, years of education, and handedness were included to undergo conventional MRI, resting-state fMRI, and high-resolution T₁WI structural image scanning.The ALFF values in six frequency bands (conventional frequency band, 0.01-0.08 Hz; slow-2, 0.198-0.25 Hz; slow-3, 0.073-0.198 Hz; slow-4, 0.027-0.073 Hz; slow-5, 0.01-0.027 Hz; and slow-6, 0-0.01 Hz) were calculated and two-sample t-test was used to compare the differences between the two groups. The receiver operating characteristic (ROC) curve was used to examine the ability of the ALFF values in multiple frequency bands to distinguish AUD from HC. Correlation analysis was used to assess the association of the parameter values of ALFF in multiple frequency bands with the clinical assessment scales. Result Thirty AUD patients and 26 HC were enrolled after head motion correction. Compared to the HC group, decreased ALFF values were detected in the four frequency bands (conventional frequency band, slow3, slow4 and slow5) in the AUD group , and the areas of decreased ALFF values were mainly located in the right superior frontal gyrus and middle frontal gyrus (FWE-corrected, voxel P<0.001, cluster P<0.05). There were no significant differences in the ALFF values for both the slow-2 band and the slow-6 band between the two groups. ROC analysis showed that ALFF values for the above differential brain regions showed good area under the curve values (range 0.829-0.873),as well as good sensitivities (range 61.5%-100.0%) and specificities (range 63.3%-96.7%). There was no significant correlation between clinical assessment scales and the ALFF values of the aforementioned differential brain regions (P>0.05). Conclusion Alterations in spontaneous brain activity in AUD patients are frequency-specific.Research on spontaneous brain activity in AUD should conduct refined frequency band analysis,and different sub-bands can provide complementary imaging information for clinical detection.

Objective This study aims to investigate the clinical features and differences in cranial MRI findings among patients with hepatolenticular degeneration, also referred to as Wilson disease (WD), complicated by epilepsy, with the purpose of providing evidence for the early clinical identification of epilepsy risk. Method sA total of 272 WD patients admitted to the Department of Encephalopathy, the First Affiliated Hospital of Anhui University of Chinese Medicine from July 2019 to August 2024 were selected. According to the discharge diagnosis, they were divided into without epilepsy associated with WD group(110 cases) and the epilepsy group (162 cases). General data, clinical types, incidence of lesions in various cerebral regions (lentiform nucleus, brainstem, thalamus, cerebral peduncle, frontal lobe, temporal lobe, parietal lobe, occipital lobe), deepening of cerebral sulci and fissures, and ventricular system dilation were compared between the two groups. Result sThe age of patients in the epilepsy group [(32.86±8.25)years] was significantly older than that in the without epilepsy associated with WD group [(30.18±11.13) years], the disease duration was [(5.86±2.45)years], which was longer than [(4.25±2.13) years](P<0.05). In terms of clinical types, the proportion of hepatotype patients in the epilepsy group (25.93%) was lower than that in the without epilepsy associated with WD group (85.45%), while the proportions of encephalotype (16.05%) was higher than those in the non-epilepsy group(5.45%) (P<0.05). Cranial MRI manifestations showed that the incidence of lesions in the lentiform nucleus, brainstem, thalamus, frontal lobe, parietal lobe, cerebral peduncle, deepening of cerebral sulci , and ventricular system dilation in the epilepsy group were significantly higher than those in the non-epilepsy associated with WD group (P<0.05).Among them, ventricular system dilation (91.98%), deepening of cerebral sulci and fissures (87.04%), and brainstem lesions (77.78%) ranked the top three in the epilepsy group. Conclusion The risk of epilepsy in WD patients is closely related to age, clinical type, and intracerebral lesions. Older patients, those with encephalotype or hepatocerebral type, and those with multiple cerebral lesions or cerebral atrophy (ventricular dilation, deepening of cerebral sulci and fissures) on cranial MRI are more likely to develop epilepsy. Clinical monitoring and intervention should be strengthened accordingly.

Objective To develop and validate an ant colony optimization-neural network (ACO-NN) prediction model based on MRI radiomics derived from pre-induction chemotherapy (IC) images to enable individualized prediction of IC response in patients with locally advanced nasopharyngeal carcinoma (LANPC). Method sData from 412 patients with pathologically confirmed LANPC treated between January 2015 and December 2023 were retrospectively collected and randomly assigned to a training set(n=289) and a validation set (n=123) at a ratio of 7∶3 .MRI radiomics features were extracted and screened using univariate rank-sum tests, Spearman correlation analysis, and elastic-net Logistic regression to identify key features. Clinically independent predictors were identified through univariate and multivariate Logistic regression analyses. Logistic regression and ACO-NN models were constructed separately, and a combined model was further developed by integrating clinically independent predictors. Model performance was evaluated using the area under the receiver operating characteristic curve (AUC), calibration curves, and decision curve analysis (DCA). Feature contributions were analyzed using Shapley additive explanations (SHAP). Result sGender, T stage, neutrophil-to-lymphocyte ratio (NLR), and albumin were identified as clinically independent predictors of IC efficacy (P<0.05).A total of 13 radiomics features with the highest predictive value were selected. The ACO-NN radiomics model demonstrated superior predictive performance to the traditional Logistic radiomics model (training set: 0.823 vs. 0.785; validation set:0.784 vs. 0.730). After integrating clinical variables with radiomics features, the ACO-NN combined model demonstrated superior predictive performance compared with the Logistic regression combined model (training set: 0.841 vs. 0.803; validation set: 0.805 vs. 0.734). Calibration curves and DCA demonstrated good agreement and clinical net benefit for the ACO-NN combined model. SHAP analysis revealed the direction and magnitude of each radiomic feature's contribution to model predictions. Conclusion The MRI-based, radiomics-enhanced ACO-NN model not only accurately predicts IC response in LANPC but also improves model interpretability, offering a noninvasive and precise tool for personalized treatment decisions.

Objective To evaluate the performance of a CT-based habitat radiomics model combined with clinical factors for predicting short-term response to transarterial chemoembolization (TACE) in patients with intermediate-to-advanced hepatocellular carcinoma(HCC). Method sThis retrospective study included 147 patients with intermediate-to-advanced HCC who underwent preoperative contrast-enhanced CT and were randomly divided into a training set(n=102) and a testing set(n=45). Tumor habitats were identified using K-means clustering, with optimal cluster number determined by the silhouette coefficient and sum of squared errors, followed by whole-cohort segmentation using a Gaussian mixture model. Radiomics features were extracted from each habitat subregion and selected using intraclass correlation coefficient analysis,Spearman correlation,maximum relevance minimum redundancy,and least absolute shrinkage and selection operator regression. Habitat-based radiomics models and a fusion model incorporating independent clinical predictors were constructed. Model performance was evaluated using receiver operating characteristic curves, calibration curves, and decision curve analysis. Result sAlpha-fetoprotein, alanine aminotransferase, and platelet count were independent clinical predictors of TACE response. The habitat-combined radiomics model achieved AUCs of 0.865 and 0.802 in the training and testing sets, respectively. After integration with clinical factors, the fusion model showed improved performance, with AUCs of 0.882 and 0.814. Decision curve and calibration analyses demonstrated good clinical utility and agreement. Conclusion A CT-based habitat radiomics model combined with clinical factors enables effective prediction of short-term response to TACE in patients with intermediate-to-advanced HCC and may assist individualized treatment decision-making.

Objective To investigate the predictive value of a model that integrates baseline intravoxel incoherent motion(IVIM) diffusion-weighted imaging(DWI) parameters with clinical and immuno-inflammatory factors for predicting transarterial chemoembolization(TACE) resistance in patients with hepatocellular carcinoma(HCC). Method sPatients with HCC were enrolled and randomly divided into a training cohort and a validation cohort at a 7∶3 ratio.Based on the occurrence of TACE resistance,patients were classified into the resistance group and the non-resistance group.Bidirectional stepwise Logistic regression analysis was used to identify independent risk factors for TACE resistance and to construct a predictive model.Model performance was assessed using receiver operating characteristic curves,calibration curves,and metrics including accuracy,recall,precision,and the F1 score. Result sA total of 194 patients with HCC were included(median age,62 years;interquartile range:59-66),of whom 79 patients(40.72%)developed TACE resistance.Compared with the non-resistance group,the resistance group had a higher proportion of patients with Child-Pugh class B,China Liver Cancer(CNLC) stage Ⅲa,and an incomplete tumor capsule;a larger tumor diameter;and higher levels of platelet count,neutrophil-to-lymphocyte ratio(NLR),and perfusion-related diffusion coefficient(D^*),along with lower apparent diffusion coefficient(ADC) values(all P<0.05).Multivariable Logistic regression analysis identified CNLC stage Ⅲa(OR=3.975,95%CI:1.198-13.185,P=0.024),elevated NLR(OR=2.211,95%CI:1.581-3.093,P<0.001),incomplete tumor capsule (OR=10.070,95%CI:3.141-32.284,P<0.001),and elevated D^*(OR=1.121,95%CI:1.024-1.227,P=0.013) as independent risk factors for TACE resistance,whereas elevated ADC was identified as a protective factor(OR=0.424,95%CI:0.226-0.793,P=0.007).The predictive model constructed using these factors yielded an area under the ROC curve of 0.932(95%CI:0.888-0.975) in the training cohort,with an accuracy of 0.896,recall of 0.857,precision of 0.889,and F1 score of 0.873.Corresponding values in the validation cohort were 0.925(95%CI:0.861-0.990),0.847,0.783,0.818,and 0.800,respectively.Calibration curves for both cohorts showed good agreement with the ideal curve. Conclusion The predictive model,incorporating CNLC stage,NLR,tumor capsule status,ADC,and D^*,exhibits good discrimination and calibration.It may serve as a reliable tool for the early identification of TACE resistance in patients with HCC.

Objective To construct a predictive model combining MRI radiomics and clinical based on machine learning,and to explore its diagnostic value in clinically significant prostate cancer(csPCa). Method sA total of 180 patients who underwent prostate MRI were retrospectively included as study subjects,including 82 patients with csPCa and 98 patients without csPCa.They were randomly divided into a training group(n=125) and a test group(n=55) at a ratio of 7∶3.Diffusion-weighted imaging,apparent diffusion coefficient,and T₂-weighted imaging(T₂WI) images were segmented and feature extraction was performed.An imaging radiomics model was constructed through cascade feature selection;independent predictors of clinical data were screened,and clinical models and combined radiomics-clinical models were constructed,respectively.The performance of each model was evaluated using receiver operating characteristic curve(ROC)and area under the curve(AUC),and the net benefit of the model was assessed using decision curve analysis(DCA);a nomogram was drawn for the optimal model. Result sThirteen radiomics features most closely related to csPCa and two independent predictors were finally selected.ROC and AUC indicated that the combined model had the best diagnostic performance,with AUC values of 0.963 and 0.927 in the training and testing groups,respectively.DCA suggested that the combined model had a higher net benefit than other models;a nomogram was also drawn for the combined model. Conclusion The MRI radiomics combined with clinical model constructed based on machine learning has high diagnostic value for csPCa.

Objective To investigate the imaging characteristics of ovarian Brenner tumors and to improve the differential diagnostic accuracy between benign and borderline/malignant forms. Method sA retrospective analysis was conducted on the clinical and imaging data of patients with pathologically confirmed Brenner tumors from five medical centers.Univariate and multivariate Logistic regression analyses were employed to identify independent risk factors associated with borderline/malignant Brenner tumors. Result sClinically, the incidence of elevated CA125 levels was significantly higher in the borderline/malignant group compared to the benign group（P<0.05）.Imaging analysis demonstrated that variables including maximum tumor diameter,tumor composition,tumor margin definition,presence of papillary projections,calcification type,degree of contrast enhancement,and CT attenuation values during arterial and venous phases were significantly different between benign and malignant tumors（P<0.05）.Multivariate Logistic regression identified maximum tumor diameter,presence of papillary projections,and CT attenuation during the venous phase as independent predictors of borderline/malignant Brenner tumors. Conclusion Larger tumor size,presence of papillary projections,and increased CT attenuation in the venous phase are indicative of a higher likelihood of borderline or malignant ovarian Brenner tumors and should be carefully considered in clinical evaluation.

Objective To assess the feasibility of a dual-energy CT（DECT） virtual non-calcium（VNCa） and clinical indicator-based random forest model for predicting R-ISS stage in multiple myeloma（MM） patients. Method sA retrospective analysis was conducted on 81 newly diagnosed and pathologically confirmed MM patients.The cohort included 52 patients with R-ISS stages Ⅰ+Ⅱ and 29 patients with stage Ⅲ. Differences in general clinical data,laboratory results,and characteristic imaging parameters were analyzed between the two groups.Based on clinical indicators,imaging features,and their combination,random forest machine learning algorithms were used to construct three predictive models:a clinical model,an imaging model,and a combined clinical-imaging model.Additionally,the SHAP algorithm was employed to enhance model transparency and clinical interpretability. Result sThe combined clinical-imaging model demonstrated the optimal diagnostic performance for predicting R-ISS,achieving an area under the curve（AUC） of 0.891.This was higher than the performance of imaging model（AUC=0.750） and clinical model（AUC=0.697）.SHAP analysis revealed that the most contributing features in the combined model,in descending order of importance,were:CT classification of bone marrow infiltration,neutrophil-to-lymphocyte ratio,lymphocyte-to-monocyte ratio,M-protein-abs,VNCa-avg,total protein,hemoglobin,CT stage of bone destruction,red blood cell count,and platelet count. Conclusion The combined model,based on DECT VNCa bone marrow imaging features and clinical indicators,provides an effective and non-invasive auxiliary assessment method for predicting R-ISS in MM patients.

Objective To observe the value of MRI meconium pattern for assessing the fetal colon and rectal development during the second and third trimesters. Method sA total of 197 healthy fetuses were analyzed retrospectively,who underwent fetal MRI examinations for reasons other than abdominal issues.The gestational weeks ranged from 21 to 37 weeks,with an average of (28.8±3.8) weeks.The T₁WI,T₂WI signals and the distribution characteristics of meconium in the colon and rectum of fetuses were analyzed.The diameters of the widest parts of the fetal meconium in the ascending colon,transverse colon,descending colon,sigmoid colon and rectum were measured respectively on the fetal coronal T₁WI sequence.The correlation with gestational weeks was analyzed using Pearson or Spearman correlation coefficient. Result sThe meconium in the fetal colon and rectum exhibited high signal intensity on T₁ weighted imaging and low signal intensity on T₂ weighted imaging.The fetal meconium accumulated gradually from distal to proximal colon with gestational.The widths of meconium in the fetal ascending colon,transverse colon,descending colon,sigmoid colon,and rectum were positively correlated with gestational age with gestational age(r=0.497,0.650,0.706,0.635,and 0.450,respectively,all P<0.001). Conclusion The distribution and width of meconium in the colon and rectum of the fetus change regularly with increasing gestational age during the second and third trimesters,offering significant value for assessing the development of the fetal colon and rectum.

Objective To develop and validate a diagnostic model for Henoch-Schönlein purpura nephritis（HSPN）in children based on multi-parametric diffusion MRI and radiomics,aiming to provide a non-invasive and precise method for the early diagnosis of HSPN. Method sChildren with Henoch-Schönlein purpura suspected of having HSPN were prospectively enrolled. All participants underwent multi-parametric diffusion MRI of the kidneys,including diffusion kurtosis imaging（DKI）and intravoxel incoherent motion（IVIM） imaging.All children were diagnosed with HSPN or non-HSPN by renal biopsy.The dataset was randomly divided into a training set and a validation set at a 7∶3 ratio.Regions of interest in the kidney were delineated on the original DKI and IVIM images.Radiomic features were extracted using variance filtering and Z-score normalization.Univariate Logistic regression and LASSO regression were employed to select the optimal set of radiomic features.Seven machine learning algorithms （gaussian process,gradient boosting decision tree,Logistic regression,quadratic discriminant analysis,random forest,support vector machine,XGBoosting） were used to construct eight radiomic models based on DKI_D,DKI_K,IVIM_D,IVIM_D^*,IVIM_f,DKI,IVIM,and DKI+IVIM,respectively.The AUC values of the models were compared to identify the optimal radiomic model.Model performance was evaluated using the receiver operating characteristic（ROC）curve and decision curve analysis（DCA）. Result sA total of 182 children with Henoch-Schönlein purpura were included in this study,comprising 127 cases in the training set （67 males,60 females;age range 6-18 years,mean age 11.34±2.53 years） and 55 cases in the validation set（27 males,28 females;age range 6-18 years,mean age 11.64±2.83 years）.The models constructed using the quadratic discriminant analysis（QDA）machine learning algorithm demonstrated the optimal performance. Among the eight radiomic models built based on this algorithm,the DKI+IVIM model performed best,with AUC values of 0.852（95%CI：0.783-0.898） in the training set and 0.909（95%CI：0.830-0.921）in the validation set.The DCA curve indicated good clinical applicability of this model. Conclusion The radiomics model based on fused DKI and IVIM images exhibits optimal diagnostic efficacy and holds promise for achieving non-invasive and accurate diagnosis of HSPN.

Objective To evaluate the value of combining conventional multiparametric enhanced MRI features with radiomics signatures and machine learning algorithms in predicting the Ki-67 index in diffuse large B-cell lymphoma (DLBCL). Method sA total of 320 pathologically confirmed DLBCL patients were retrospectively enrolled,including 170 with Ki-67<75% and 150 with Ki-67≥75%.They were divided into training and testing sets at a 6∶4 ratio.Regions of interest(ROIs)were manually delineated on preoperative enhanced MRI images(T₁WI, T₂WI,contrast-enhanced T₁WI).Conventional imaging features(such as T₁WI signal,necrosis,multi-tissue involvement,enhancement pattern,ADCmean)and radiomics features(Delta Radscore,T₂ Radscore)were extracted.Using the training set,multiple machine learning models(SVM,KNN,XGBoost,etc.)and Logistic regression were employed to develop prediction models,which were then validated on the testing set.Model performance was assessed using DeLong’s test,decision curve analysis(DCA),SHAP plots,nomograms,and calibration curves. Result sMultivariate analysis identified delta radiomics score,T₂ radiomics score,T₁WI signal,ADCmean,and necrosis as independent factors for Ki-67 classification in DLBCL.The XGBoost model demonstrated the best discriminative performance(training AUC 0.98,testing AUC 0.88).DCA indicated high clinical net benefit for the XGBoost model.SHAP plots highlighted delta radiomics score,T₂ radiomics score,T₁WI signal,and ADCmean as key predictors.A nomogram constructed based on these features showed good calibration. Conclusion Integrating conventional multiparametric enhanced MRI features,particularly T₁WI signal,with radiomics signatures(Delta Radscore),and applying the XGBoost machine learning algorithm can effectively and efficiently predict the Ki-67 index in DLBCL patients prior to radiochemotherapy.The nomogram based on key features provides a valuable auxiliary tool for clinical decision-making.

Objective To investigate the clinical value of a personalized acquisition window optimization scheme based on heart rate stratification,combined with Deep Learning Image Reconstruction and the second-generation Snapshot Freeze technology,in ECG-less Coronary Computed Tomography Angiography. Method sA total of 160 patients undergoing CCTA were prospectively and consecutively enrolled and randomly divided into a fixed wide window group(Group A,n=80) and a heart rate stratified personalized window group(Group B,n=80).Group A utilized a fixed 600 ms acquisition window,while Group B adopted personalized narrow acquisition windows based on heart rate(≤60,60-75,75-90,and >90 bpm).Both groups applied DLIR-H and SSF2 algorithms for image post-processing.The aortic root SD,CT values of coronary artery branches(RCA,LAD,LCX),signal-to-noise ratio(SNR),contrast-to-noise ratio(CNR),subjective image quality scores,and radiation dose metrics(DLP,CTDIvol) were compared between the two groups. Result sThere were no statistically significant differences in general characteristics between the two groups(P>0.05).In terms of Objective image quality,no significant differences were observed between Group A and Group B regarding SDaorta and CT values of coronary branches(P>0.05).Inter-observer agreement was good(Kappa=0.692).There was no significant difference in subjective image quality scores between the two groups(P>0.05),and all images met clinical diagnostic requirements.However,the DLP in Group B was significantly lower than that in Group A across all heart rate subgroups:reduced by 25.8% in the ≤60 bpm subgroup(113.03±7.82 vs. 152.41±21.29),15.2% in the 60-75 bpm subgroup(130.89±12.53 vs. 154.30±17.97),17.6% in the 75-90 bpm subgroup(125.67±16.95 vs. 152.59±18.17),and 20.8% in the >90 bpm subgroup(121.45±11.41 vs. 153.25±16.83)(all P<0.05). Conclusion The personalized acquisition window optimization scheme based on heart rate stratification,combined with DLIR and SSF2 technologies,can reduce radiation dose by 15.2% to 25.8% in patients with different heart rates while maintaining image quality in ECG-less CCTA.This approach provides a safer and more effective personalized CCTA scanning protocol,achieving the clinical goal of low-dose and high-quality coronary imaging.

Objective To investigate the feasibility of using photon-counting detector CT(PCD-CT)with low keV combined with quantum iterative reconstruction(QIR)in dual-phase contrast-enhanced CT images of esophageal cancer,and to determine whether higher QIR levels yield better image quality.Compare the optimal QIR level with the image quality of energy-integrating detector CT(SOMATOM Force). Method sThis study prospectively collected arterial and venous phase images from 86 patients clinically diagnosed with esophageal cancer at the First Affiliated Hospital of Zhengzhou University between April 2024 and November 2025,using photon-counting CT.Axial images were reconstructed at 50 keV with iterative reconstruction algorithms(QIR-1 to QIR-4)and QIR-off.Additionally,dual-phase contrast-enhanced CT images were prospectively collected from 92 patients clinically diagnosed with esophageal cancer during the same period using energy-integrating detector CT(EID-CT).For Objective image quality evaluation,the signal-to-noise ratio(SNR)and contrast-to-noise ratio(CNR)of the lesions and muscles were calculated.For subjective image quality evaluation,a 5-point scale was used to score lesion visibility,diagnostic confidence,and image noise for each PCD-CT group.Comparisons of SNR and CNR among the five groups were performed using one-way analysis of variance（ANOVA).Comparisons of image quality scores were conducted using the Friedman test,with post-hoc pairwise comparisons adjusted using Bonferroni correction. Result sThe overall differences in both subjective and Objective scores were statistically significant(P<0.001).In the Objective evaluation of image quality,comparisons were made between images reconstructed with QIR-off and QIR-1-3 versus QIR-4.For lesion CNR in both arterial and venous phase images,QIR-4 was significantly higher than QIR-off and QIR-1 images(P<0.05),but showed no statistically significant difference compared to QIR-2 and QIR-3 images(P_venous=0.243,1.000;P_arterial=0.651,1.000).For lesion SNR and muscle SNR in both arterial and venous phase images,QIR-4 was significantly higher than QIR-off and QIR-1-3 images(P<0.05).For image noise（SD）in both arterial and venous phase images,QIR-4 was significantly lower than QIR-off and QIR-1-3 images（P<0.05).For subjective noise and diagnostic confidence scores in both arterial and venous phase images,QIR-4 was significantly higher than QIR-off and QIR-1-3 images（P<0.05).For lesion visibility,there was no statistically significant difference between QIR-3 and QIR-4 groups in either arterial or venous phase（P_venous=0.847;P_arterial=0.341),while QIR-4 was significantly higher than QIR-off and QIR-1-2 images（P<0.05).Comparison of subjective and Objective evaluation indicators between EID-CT images and QIR-4 images:For lesion CNR and lesion SNR in both arterial and venous phase images,QIR-4 was significantly higher than EID-CT（P<0.05).For muscle SNR and image noise（SD),QIR-4 images were comparable to EID-CT,with no statistically significant differences（P>0.05).For subjective noise scores in both arterial and venous phase images,QIR-4 was comparable to EID-CT images,with no statistically significant differences（P_venous=0.762;P_arterial=0.673).For lesion visibility and diagnostic confidence,QIR-4 was superior to EID-CT images in both arterial and venous phases,with statistically significant differences（P<0.001). Conclusion The tissue contrast and spatial resolution of QIR-4 images for esophageal cancer were superior to those of QIR-off and QIR-1 images,and comparable to those of QIR-3 images.Additionally,QIR-4 images significantly reduced image noise compared with the other reconstruction groups.Therefore,QIR-4 was determined to be the optimal reconstruction level.The lesion visibility of QIR-4 images was superior to that of EID-CT,while the image noise was comparable to that of EID-CT images.

Objective To compare body fat distribution patterns among patients with maturity-onset diabetes of the young type 2(MODY2),type 2 diabetes mellitus(T2DM),and non-diabetic controls using MRI proton density fat fraction（PDFF). Method sPatients who underwent abdominal MRI examinations at Peking University People’s Hospital from April 2017 to May 2025 were collected from the Radiology Information System(RIS).They were divided into three groups based on clinical diagnosis:MODY2,T2DM,and non-diabetic controls.The latter two groups were matched with the MODY2 group in terms of sex,age,and body mass index(BMI),while excluding patients with conditions such as malignancies,autoimmune diseases,and diffuse liver or pancreatic diseases.PDFF values of the liver,pancreas,and other regions were measured using MRI-PDFF sequences. Result sA total of 147 patients were collected,including MODY2(n=18),T2DM(n=21),and control group（n=108).The median liver PDFF values for the three groups were 2.3%（1.5%,3.9%）,3.6%（2.7%,9.9%）,and 3.2%（2.3%,5.0%）,respectively,while the median pancreatic PDFF values were 2.4%（1.6%,3.4%）,3.9%（3.0%,4.7%）,and 3.2%（2.5%,4.4%）,respectively.The liver PDFF in the MODY2 group was lower than that in the T2DM group（P=0.015),and the pancreas PDFF was lower than that in the T2DM group（P=0.002）and the control group（P=0.016).Liver PDFF showed a moderate positive correlation with blood triglycerides（rₛ=0.475,P<0.001). Conclusion sMODY2 patients exhibit a distinct body fat distribution pattern:lower hepatic and pancreatic fat content,with liver fat content positively correlating with blood triglyceride levels.The MRI-PDFF sequence can effectively differentiate differences in organ fat content among the three groups.

Objective To compare the occurrence of double-contour artifacts in contrast-enhanced mammography(CEM)between two generations of GE Mammography machine(Senographe Essential 2019 and Senographe Pristina Hygeia 2023),analyze the artifact control effectiveness and technical advantages of the new model,and provide a basis for optimizing clinical CEM equipment application and image quality management. Method sImaging data of female patients who underwent CEM at a single institution were retrospectively reviewed.Patients were allocated into two groups according to the equipment used:the older system group(Senographe Essential) and the newer system group(Senographe Pristina Hygeia),with 200 patients in each group.All patients underwent standard CEM in craniocaudal(CC) and mediolateral oblique(MLO) projections.A total of 800 subtracted images were acquired per group,and 1600 images in total were included in the analysis.A 4‑point grading scale for double‑contour artifacts was applied.Two experienced breast radiologists independently interpreted images in a double‑blind fashion.Interobserver agreement was assessed using the Kappa test.Differences in artifact incidence between groups were analyzed using Pearson χ² test to evaluate the effect of hardware and technical upgrades on artifact reduction. Result sThe baseline data of the two groups were balanced and comparable(all P>0.05).The interobserver agreement between the two radiologists was good(Kappa=0.83,P<0.001).The total incidence of double-contour artifacts in the new group was 9.50%,significantly lower than that in the old group(61.25%)(χ²=468.58,P<0.001).The incidence of moderate-to-severe artifacts in the new group was 3.12%,significantly lower than that in the old group(55.37%)(χ²=527.692,P<0.001). Conclusion The new generation of GE Mammography machine significantly reduces the incidence and severity of CEM double-contour artifacts through Hardware and software technological improvements, such as rapid high-low energy exposure switching,intelligent target/filter matching,high-sensitivity detectors,high grid ratio,and an optimized subtraction algorithms,enhancing image stability and diagnostic reliability,with important clinical application value.

Objective To investigate the feasibility of quantitative T₂ mapping values in assessing inflammatory activity and intestinal fibrosis in a rat model of ulcerative colitis（UC）,and to evaluate the value of T₂ mapping as a noninvasive and quantitative method for assessing UC inflammatory injury. Method sA total of 36 SPF-grade Sprague-Dawley rats were randomly divided into three groups（n=12 each）：normal group,acute UC group,and chronic UC group.Acute and chronic UC models were established by administering dextran sodium sulfate（DSS）in drinking water,while the normal group received pure water.After successful modeling,all rats underwent MRI T₂ mapping to obtain quantitative T₂ values of the colon.Serum levels of interleukin-6（IL-6）,lipopolysaccharide（LPS）,and tumor necrosis factor-alpha（TNF-α）were measured by enzyme-linked immunosorbent assay.Hematoxylin-eosin（HE）staining was used to determine histopathological injury scores,and Picrosirius Red stain was used to assess the degree of fibrosis.One-way analysis of variance,Spearman correlation analysis,and stepwise multiple linear regression analysis were performed for statistical analysis. Result sThe differences in T₂ values among the three groups of rats were statistically significant,with T₂ values in both the acute and chronic groups being higher than those in the normal control group（P<0.001）.The trends of IL-6,LPS,and TNF-α levels across the three groups were highly consistent with the T₂ value changes,showing that the acute and chronic groups had higher levels than the normal group（all P<0.001）,and the TNF-α level in the acute group was higher than that in the chronic group（P<0.05）.HE staining showed that the pathological injury scores in the acute and chronic groups were higher than those in the normal group（P<0.05）,and the chronic group had a higher score than the acute group（P<0.05）.The difference in fibrosis area fraction among the three groups was statistically significant（P<0.05）,with the chronic group showing significantly higher fibrosis levels than the normal group（P<0.05）.T₂ values were significantly positively correlated with IL-6,LPS,TNF-α levels,and HE pathological injury scores（P<0.01）,but the correlation with fibrosis area fraction did not reach statistical significance（P=0.254）.Stepwise multiple linear regression analysis indicated that acute group,chronic group and fibrosis area fraction were independent influencing factors for T₂ values. Conclusion T₂ mapping can serve as a potential imaging biomarker for non-invasive and quantitative assessment of intestinal inflammation,edema,and inflammatory activity in the DSS-induced UC models.T₂ values in the UC rat intestine are closely related to inflammatory cytokines,but the ability of T₂ mapping to identify intestinal fibrosis is influenced by factors such as the degree of inflammation,and it is not suitable for assessing intestinal fibrosis alone.This study is limited by factors such as sample size,DSS modeling method,and motion artifacts during scanning.Future studies should involve larger sample sizes,optimized modeling protocols,additional dynamic monitoring,and integration of multiparametric MRI and radiomics approaches for further improvement.