Background and purpose:Diffuse large B-cell lymphoma(DLBCL)is subclassified into germinal center B-cell-like(GCB)and non-GCB subtypes,which differ in prognosis and treatment response.However,current distinction still relies on invasive pathological assays.This study developed radiomics and deep-learning models based on multiparametric magnetic resonance imaging(MRI)to non-invasively differentiate the two subtypes preoperatively,thereby reducing dependence on histopathological examination.Methods:This study retrospectively included patients with pathologically confirmed DLBCL diagnosed at Huashan Hospital,Fudan University,and other institutions between March 2013 and December 2024.Using multiparametric MRI data,we developed DLBCL-subtype classification models that combined 4 radiomics-based machine-learning algorithms:support vector machine(SVM),logistic regression(LR),Gaussian process(GP)and Naive Bayes(NB),with 3 deep-learning architectures[densely-connected convolutional networks 121(DenseNet121),residual network 101(ResNet101)and EfficientNet-b5].Additionally,two radiologists with different experience levels independently classified DLBCL on MRI in a blinded fashion.Model and radiologist performance were quantified using the area under the receiver operating characteristic curve(AUC),accuracy(ACC),and F1-score to evaluate their ability to distinguish GCB from non-GCB subtypes.This study was approved by the Ethics Committee of Huashan Hospital of Fudan University(No.KY2024-663),and all patients signed informed consents.Results:A total of 173 patients were enrolled(55 with GCB subtype and 118 with non-GCB subtype).Radiomics and deep learning methods effectively distinguished DLBCL subtypes.Among these,the GP radiomics model(based on T1-CE+T2-FLAIR+ADC sequences)and DenseNet121 deep learning model(based on T1-CE+T2-FLAIR+ADC sequences)demonstrated optimal performance.Both achieved excellent results on the internal validation set(GP:AUC=0.900,ACC=0.896,F1=0.840;DenseNet121:AUC=0.846,ACC=0.854,F1=0.774)and maintained robustness on the external validation set.Furthermore,the classification efficacy of the optimal AI model surpassed that of experienced radiologists(highest physician AUC=0.678).Conclusion:Radiomics and deep-learning models based on multiparametric MRI features can effectively differentiate GCB from non-GCB subtypes of DLBCL.Among them,GP and DenseNet121 exhibit outstanding performance,especially when integrating multi-sequence feature sets for classifying DLBCL subtypes on complex imaging data.

Background and purpose:Diffuse large B-cell lymphoma(DLBCL)is subclassified into germinal center B-cell-like(GCB)and non-GCB subtypes,which differ in prognosis and treatment response.However,current distinction still relies on invasive pathological assays.This study developed radiomics and deep-learning models based on multiparametric magnetic resonance imaging(MRI)to non-invasively differentiate the two subtypes preoperatively,thereby reducing dependence on histopathological examination.Methods:This study retrospectively included patients with pathologically confirmed DLBCL diagnosed at Huashan Hospital,Fudan University,and other institutions between March 2013 and December 2024.Using multiparametric MRI data,we developed DLBCL-subtype classification models that combined 4 radiomics-based machine-learning algorithms:support vector machine(SVM),logistic regression(LR),Gaussian process(GP)and Naive Bayes(NB),with 3 deep-learning architectures[densely-connected convolutional networks 121(DenseNet121),residual network 101(ResNet101)and EfficientNet-b5].Additionally,two radiologists with different experience levels independently classified DLBCL on MRI in a blinded fashion.Model and radiologist performance were quantified using the area under the receiver operating characteristic curve(AUC),accuracy(ACC),and F1-score to evaluate their ability to distinguish GCB from non-GCB subtypes.This study was approved by the Ethics Committee of Huashan Hospital of Fudan University(No.KY2024-663),and all patients signed informed consents.Results:A total of 173 patients were enrolled(55 with GCB subtype and 118 with non-GCB subtype).Radiomics and deep learning methods effectively distinguished DLBCL subtypes.Among these,the GP radiomics model(based on T1-CE+T2-FLAIR+ADC sequences)and DenseNet121 deep learning model(based on T1-CE+T2-FLAIR+ADC sequences)demonstrated optimal performance.Both achieved excellent results on the internal validation set(GP:AUC=0.900,ACC=0.896,F1=0.840;DenseNet121:AUC=0.846,ACC=0.854,F1=0.774)and maintained robustness on the external validation set.Furthermore,the classification efficacy of the optimal AI model surpassed that of experienced radiologists(highest physician AUC=0.678).Conclusion:Radiomics and deep-learning models based on multiparametric MRI features can effectively differentiate GCB from non-GCB subtypes of DLBCL.Among them,GP and DenseNet121 exhibit outstanding performance,especially when integrating multi-sequence feature sets for classifying DLBCL subtypes on complex imaging data.

Objective To investigate the prevalence of low back pain among adults in Beijing.Methods The study design was a cross-sectional study,and the multi-stage sampling was used.A questionnaire survey was conducted in December 2010 to investigate prevalence of low back pain in adults who had lived in Beijing for over 6 months.Total prevalence and prevalence by region,gender and age were calculated.The chi-square test was used to compare results.Results A total of 3860 people were enrolled in this study.The one-year prevalence of low back pain was 26.09％ (1007/3860),and the point prevalence was 6.11％ (236/3860).The prevalence of different duration of low back pain (3 months,3-6 months,≥6 months) was 16.76％ (647/3860),4.12％ (159/3860) and 5.21％ (201/3860),respectively.The prevalence of females (28.83％) was higher than that of males (23.03％).The prevalence among different regions was significantly different.Prevalence in suburb and rural area (29.88％ and 27.54％,respectively) was higher than that in urban area (20.88％).No matter males or females,the prevalence in urban area was the lowest (17.48％ and 24.00％,respectively).With the increasing of age,the prevalence of low back pain became higher.In males,the prevalence of 55 to 59 years group was highest,while 60 to 64 years group was highest in females.In urban area and rural area,the prevalence of 60 to 64 years group was highest (34.43％ and 48.68％,respectively),while 55 to 59 years group was highest in suburb (47.26％).Conclusion The oneyear and point prevalence of low back pain among adults in Beijing are higher,with wide distribution.The chronic low back pain is more common.The prevalence of low back pain is higher in suburb and rural area.Females have a higher prevalence than males.Moreover,the prevalence of low back pain increases with age.

Objective To investigate the prevalence and distribution of lumbar degeneration disease among adults in Beijing.Methods The study design was cross-sectional study.The multi-stage sampling was used.The study objects were residents who were lived in Beijing over six months and older than 18 years.The related information was obtained by self-designed questionnaire.Single and multivariable Logistic regression models were applied to analyze the high risk populations.Results A total of 3186 people were studied.There were 292 people who suffered lumbar degenerational disease.The prevalence was 9.17％.The prevalence at downtown,suburb county and rural area was 7.88％,10.20％,and 9.59％,respectively,and there was not significant difference (x2=3.545,P=0.170).The prevalence of female (10.05％) was higher than male (8.13％) (x2=4.081,P=0.043; OR=1.337,95％CI:1.044,1.713).The prevalence of the population who was not less than 45 years older was significant higher than that of the population less than 45 years older (x2=102.982,P＜ 0.001).The physical labor group (12.16％) had higher risk for lumbar degeneration disease compared with mixed group (6.65％)(OR=1.510,95％CI:1.102,2.071).There were no significant differences in different education,social insurance,and income populations.Conclusion The prevalence of lumbar degeneration disease among adults is much higher and vastly distributed.It's necessary to reinforce the prevention,diagnosis and treatment study.People of female,more than 45 years older and physical labor group are high risk populations.

Objective To explore the effects of proton pump inhibitors (PPIs) therapy on esophageal acid exposure of patients with gastroesophageal reflux disease(GERD),and the correlation of anxiety and depression with recurrence of acid-related symptoms after discontinuation of PPIs.Methods From February 2010 to June 2011,28 patients with GERD diagnosed by ambulatory 24 h esophageal pH monitoring admitted to Beijing Jishuitan Hospital were treated with esomeprazole 20 mg 2 times/d for 8 weeks (male 16,female 12).Symptoms after drug discontinuation were monitored.Ambulatory 24 h esophageal pH monitoring was performed on patients,whose symptom recurred within 8 weeks after treatment.BMI,Self-rating Anxiety Scale(SAS),and Self-rating Depression Scale (SDS) were detected.Results Among the 28 patients with GERD,15 (53.6％) recurred symptoms after withdraw of PPIs.Compared with the asymptomatic group after withdraw of PPIs,the pretreatment duration of pH 4 (supine),24 h total acid reflux time,number of time periods with acid reflux ＞5 minutes,the maximal acid reflux time and 24 h total number of acid reflux in the symptomatic recurrence group were statistically significantly increased ( 11.7％vs 4.5％,138.8 minutes vs 62.1 minutes,6.0 vs 2.0,27.0 minutes vs 12.4 minutes,74.0 times vs43.0times,respectively,all P values ＜ 0.05 ).There were no significant differences in BMI,SAS and SDS between the two groups.Conclusions The basic level of esophageal acid exposure of patients with GERD before PPIs therapy may influence the esophageal acid exposure after PPIs therapy and then may affect the recurrence of symptoms.Although anxiety and depression is common in patients with GERD,it is not found that the recurrence of acid-related symptoms after the discontinuation of PPIs therapy is related to the anxiety and depression.