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 explore the effect of a self-developed augmented multi-task sensory exer-cise intervention on perceptual-motor skills and motor coordination in children with developmental coor-dination disorder(DCD).Methods Twenty-four DCD children aged 6～7 years were included and ran-domly divided into a control group(n=12)and an experimental group(n=12).Both groups maintained their daily physical activities and took P.E.classes in school,while the experimental group additional-ly underwent a multi-task somatosensory motor intervention consisting of 3 stages.Each stage lasted for 4 weeks,3 sessions per week,with progressively increased difficulty.A self-designed data process-ing module of the somatosensory intervention system was used to collect the scores of each task(i.e.,sensory-motor evaluation),and the changes in scores of each task at each stage were analyzed.More-over,before and after the intervention,the motor coordination ability was assessed using the Move-ment Assessment Battery for Children-Second Edition(MABC-2),and the intra-and inter-group dif-ferences in MABC-2 scores were analyzed.Two weeks after the intervention,the experimental group conducted 3 sessions of motor relearning at the difficulty level of the 3rd stage to examine the effects of motor retention and relearning.Results 1)The experimental group showed a significant increase in their scores on all tasks from T1 to T5 during the enhanced multi-task somatosensory training of all the three stages(P<0.001,P<0.05).2)After the kinesthetic exercises,the experimental group had sig-nificantly higher abilities in fine motor skills,positioning&grasping and body balance compared to the control group(P<0.05),with relatively greater effect on threading beads,drawing traces,tossing and catching bags,single-leg balance,tiptoe walking and two-legged hopping(η2≥0.114).3)In the assessment of motor skill retention and relearning,the T1～T4 scores in the first retest were lower than the previous final ones(P<0.05),with no significant difference from those at the end of the second stage,but showing significantly higher T1 and T5 scores.Moreover,all the T1～T5 scores reached the level measured at the previous end of stage three after three sessions of relearning.Conclusion The augmented multi-task somatosensory practice based on perception-motor skills theory can enhance the perception-motor skills and motor coordination ability of the DCD children,with good motor skill reten-tion and recovery effectiveness after such intervention.

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 value of synthetic MRI(syMRI)for evaluating structural changes of brain nuclei and white matter in patients with Parkinson disease(PD).Methods Forty-five PD patients(PD group)and 35 healthy controls(HC group)were prospectively enrolled.Head scanning with strategically acquired gradient echo sequence was performed,then syMRI was obtained with post-processing.The values of T1,proton density(PrD),R2* and magnetic sensitivity values(MSV)of nuclei and white matter were measured based on syMRI and compared between groups.The correlations of syMRI quantitative parameters of nuclei and white matter being significantly different between groups with neurological function scores of subject were analyzed,and a combined model was established using multivariate logistic regression analysis.Receiver operating characteristic(ROC)curve was drawn,the efficacy of single syMRI quantitative parameter and combined model for diagnosing PD was evaluated.Results Compared with HC group,T1 and PrD values of bilateral caudate nuclei decreased,while MSV of bilateral substatia nigra and right red nucleus increased(all FDR correction P＜0.05),while T1 and PrD values of bilateral splenium of corpus callosum,PrD values of left posterior limbs of the internal capsule(PLIC),R2* values of bilateral PLIC and right genu of the corpus callosum decreased in PD group(all FDR correction P＜0.05).syMRI quantitative parameters of the above brain regions were correlated with scores of Montreal cognitive assessment and Hamilton anxiety scale(r=-0.390-0.416,all P＜0.01).The area under the curve of single syMRI quantitative parameter for diagnosing PD was 0.664-0.788,lower than that of combined model(0.905,Z=2.653-4.096,all P＜0.05).Conclusion SyMRI could be used to quantitatively evaluate structural changes of brain nuclei and white matter in PD patients.

Objective To explore the effect of a self-developed augmented multi-task sensory exer-cise intervention on perceptual-motor skills and motor coordination in children with developmental coor-dination disorder(DCD).Methods Twenty-four DCD children aged 6～7 years were included and ran-domly divided into a control group(n=12)and an experimental group(n=12).Both groups maintained their daily physical activities and took P.E.classes in school,while the experimental group additional-ly underwent a multi-task somatosensory motor intervention consisting of 3 stages.Each stage lasted for 4 weeks,3 sessions per week,with progressively increased difficulty.A self-designed data process-ing module of the somatosensory intervention system was used to collect the scores of each task(i.e.,sensory-motor evaluation),and the changes in scores of each task at each stage were analyzed.More-over,before and after the intervention,the motor coordination ability was assessed using the Move-ment Assessment Battery for Children-Second Edition(MABC-2),and the intra-and inter-group dif-ferences in MABC-2 scores were analyzed.Two weeks after the intervention,the experimental group conducted 3 sessions of motor relearning at the difficulty level of the 3rd stage to examine the effects of motor retention and relearning.Results 1)The experimental group showed a significant increase in their scores on all tasks from T1 to T5 during the enhanced multi-task somatosensory training of all the three stages(P<0.001,P<0.05).2)After the kinesthetic exercises,the experimental group had sig-nificantly higher abilities in fine motor skills,positioning&grasping and body balance compared to the control group(P<0.05),with relatively greater effect on threading beads,drawing traces,tossing and catching bags,single-leg balance,tiptoe walking and two-legged hopping(η2≥0.114).3)In the assessment of motor skill retention and relearning,the T1～T4 scores in the first retest were lower than the previous final ones(P<0.05),with no significant difference from those at the end of the second stage,but showing significantly higher T1 and T5 scores.Moreover,all the T1～T5 scores reached the level measured at the previous end of stage three after three sessions of relearning.Conclusion The augmented multi-task somatosensory practice based on perception-motor skills theory can enhance the perception-motor skills and motor coordination ability of the DCD children,with good motor skill reten-tion and recovery effectiveness after such intervention.

Objective To investigate the value of synthetic MRI(syMRI)for evaluating structural changes of brain nuclei and white matter in patients with Parkinson disease(PD).Methods Forty-five PD patients(PD group)and 35 healthy controls(HC group)were prospectively enrolled.Head scanning with strategically acquired gradient echo sequence was performed,then syMRI was obtained with post-processing.The values of T1,proton density(PrD),R2* and magnetic sensitivity values(MSV)of nuclei and white matter were measured based on syMRI and compared between groups.The correlations of syMRI quantitative parameters of nuclei and white matter being significantly different between groups with neurological function scores of subject were analyzed,and a combined model was established using multivariate logistic regression analysis.Receiver operating characteristic(ROC)curve was drawn,the efficacy of single syMRI quantitative parameter and combined model for diagnosing PD was evaluated.Results Compared with HC group,T1 and PrD values of bilateral caudate nuclei decreased,while MSV of bilateral substatia nigra and right red nucleus increased(all FDR correction P＜0.05),while T1 and PrD values of bilateral splenium of corpus callosum,PrD values of left posterior limbs of the internal capsule(PLIC),R2* values of bilateral PLIC and right genu of the corpus callosum decreased in PD group(all FDR correction P＜0.05).syMRI quantitative parameters of the above brain regions were correlated with scores of Montreal cognitive assessment and Hamilton anxiety scale(r=-0.390-0.416,all P＜0.01).The area under the curve of single syMRI quantitative parameter for diagnosing PD was 0.664-0.788,lower than that of combined model(0.905,Z=2.653-4.096,all P＜0.05).Conclusion SyMRI could be used to quantitatively evaluate structural changes of brain nuclei and white matter in PD patients.

Objective To observe the influence of different acceleration factors(AF)on compressed sensing susceptibility weighted imaging(CS-SWI)for cerebral medullary veins of healthy people,and to screen the best AF.Methods Forty healthy volunteers were prospectively enrolled.Axial brain SWI images were obtained with CS technique under different AF(AF0,CS2,CS4,CS6,CS8 and CS10),and the phase value(PV)and standard deviation(SD)of bilateral septal vein(SV),internal cerebral vein(ICV),thalamus vein(TV),basal vein(BV)and dentate nucleus vein(DNV)were measured.Taken PV and SD of parietal white matter as controls,the signal-to-noise ratio(SNR)and contrast-to-noise ratio(CNR)of venous images were calculated.Then the original images were reconstructed with minimum intensity projection,and the subjective quality scoring of reconstructed images were performed using a 5-point scale.SNR,CNR,PV and quality score were compared among images under different AF,and the best AF,i.e.with the best performances for displaying and quantitatively analyzing cerebral medullary veins in healthy people was obtained.Results Compared with those acquired with AF0,SNR and CNR of all cerebral medullary veins acquired with CS6,CS8 and CS10 were significantly different(all adjusted P＜0.05).Meanwhile,significant differences of PV in bilateral SV and right TV were found among CS6,CS8 and CS10,also in bilateral ICV,left TV and bilateral BV between CS8 and CS10(all adjusted P＜0.05).Conclusion Excessive AF might decrease image quality of CS-SWI for cerebral medullary veins.CS4 was the best AF for displaying and quantitatively analyzing cerebral medullary veins in healthy people.

Objective:To investigate the role of HMGB1-TLR4-NF-κB signaling pathway in cerebral ischemia-reperfusion in-jury and the effect of adenosine preconditioning on the signaling pathway.Methods:Total 80 adult male Sprague-Dawley rats weighing 220～270 g were selected from the Animal Center of Xinxiang Medical University.The rats were randomly divided into F group(sham operation group),I/R group(ischemia reperfusion group)and AP group(adenosine preconditioning group).The MCAO model of rats was established by wire embolization.Quantitative analysis of neural function in successfully modeled rats using animal behavior scor-ing method,the morphological changes of brain cells were observed by HE staining,TTC staining was used to observe cerebral infarc-tion and cerebral infarction volume was calculated;Immunohistochemical staining was used to detect HMGB1,TLR4 and NF-κB pro-tein expression levels in brain tissues of each group.The data were statistically analyzed by one-way ANOVA in SPSS26.0 software.Results:After ischemia reperfusion,the neurological function of I/R group and AP group showed different degrees of impairment,and the neurological function scores of the two groups were significantly higher than that of F group,the difference was statistically signifi-cant(P<0.05),and the neurological function of the AP group was significantly less than that of I/R group,the difference was also sta-tistically significant(P<0.05).TTC staining showed that AP group,I/R group rat cerebral infarction volume was significantly more than F group[(93.670±4.509)mm3,(123.670±7.234)mm3 vs(0.000±0.000)mm3],and AP group rats infarction volume was signifi-cantly reduced than that in I/R group,the difference had statistical significance(P<0.05).Immunohistochemistry showed that HMGB1,TLR4,NF-κB protein in F group with a small amount of expressions in rats,while significantly expressed in AP group and I/R group relatively,and the AP group of each subgroup rat HMGB1,TLR4,NF-κB protein expressions significantly lower than the amount of I/R group,the difference had statistical significance(P<0.05).Conclusion:Adenosine preconditioning can reduce the expressions of HMGB1,TLR4 and NF-κB protein,and then protect the rats with cerebral ischemia-reperfusion injury.

Objective: To investigate the prognostic utility of radiomics features extracted from 18 F-fluorodeoxyglucose (FDG) PET/CT combined with clinical factors and metabolic parameters in predicting progression-free survival (PFS) and overall survival (OS) in individuals diagnosed with extranodal nasal-type NK/T cell lymphoma (ENKTCL). Materials and Methods: A total of 126 adults with ENKTCL who underwent 18 F-FDG PET/CT examination before treatment were retrospectively included and randomly divided into training (n = 88) and validation cohorts (n = 38) at a ratio of 7:3.Least absolute shrinkage and selection operation Cox regression analysis was used to select the best radiomics features and calculate each patient’s radiomics scores (RadPFS and RadOS). Kaplan–Meier curve and Log-rank test were used to compare survival between patient groups risk-stratified by the radiomics scores. Various models to predict PFS and OS were constructed, including clinical, metabolic, clinical + metabolic, and clinical + metabolic + radiomics models. The discriminative ability of each model was evaluated using Harrell’s C index. The performance of each model in predicting PFS and OS for 1-, 3-, and 5-years was evaluated using the time-dependent receiver operating characteristic (ROC) curve. Results: Kaplan–Meier curve analysis demonstrated that the radiomics scores effectively identified high- and low-risk patients (all P < 0.05). Multivariable Cox analysis showed that the Ann Arbor stage, maximum standardized uptake value (SUVmax), and RadPFS were independent risk factors associated with PFS. Further, β2-microglobulin, Eastern Cooperative Oncology Group performance status score, SUVmax, and RadOS were independent risk factors for OS. The clinical + metabolic + radiomics model exhibited the greatest discriminative ability for both PFS (Harrell’s C-index: 0.805 in the validation cohort) and OS (Harrell’s C-index: 0.833 in the validation cohort). The time-dependent ROC analysis indicated that the clinical + metabolic + radiomics model had the best predictive performance. Conclusion The PET/CT-based clinical + metabolic + radiomics model can enhance prognostication among patients with ENKTCL and may be a non-invasive and efficient risk stratification tool for clinical practice.

Objective To investigate the cancer death and distribution characteristics of residents in Cixian County. Methods In accordance with the norms of cancer registration, cancer death data from 2013 to 2017 in Cixian were collected and analyzed, and the crude cancer death rate, age-standardized mortality rates by the Chinese standard population (ASMRC), age-standardized mortality rates by the global standard population (ASMRW). Results From January 1st, 2013, to December 31st, 2017, 6 490 cases of cancer death were recorded. The average annual crude mortality rate was 202.88/100 000, ASMRC was 186.49/100 000, and the ASMRW was 189.02/100 000. The top 10 male mortality cancers were esophageal cancer, stomach cancer, trachea, bronchus and lung cancer, liver cancer, rectal cancer, cerebral nervous system cancer, colon cancer, leukemia, pancreatic cancer, and bladder cancer in order. The top 10 female mortality cancers were esophageal cancer, trachea, bronchus and lung cancer, stomach cancer, liver cancer, breast cancer, cervical cancer, colon cancer, brain, nervous system cancer, rectal cancer, and ovarian cancer. The age of death increased considerably from the age of 40 years. It increased with the increase in age and reached the peak at the age of 85 years. Conclusion Upper gastrointestinal cancer and lung cancer were the main cancers that threatened the residents of Cixian County from 2013 to 2017. Screening and comprehensive prevention of high-risk groups are still the main targets of cancer prevention and control.