Objective:To analyze the influencing factors of postoperative pulmonary embolism (PE) in patients undergoing hepatectomy.Methods:Clinical data of 171 patients undergoing hepatectomy at the Department of Hepatobiliary and Pancreatic Surgery, the Second Affiliated Hospital of Kunming Medical University from January 2018 to November 2024 were retrospectively analyzed, including 95 males and 76 females, aged (52.0±12.6) years. The pathology of patients included hepatolithiasis, hepatic hemangioma, liver abscess, focal nodular hyperplasia of liver, hepatocellular carcinoma, intrahepatic cholangiocarcinoma, etc. According to whether PE occurred after hepatectomy, patients were divided into PE group ( n=64) and control group ( n=107). Univariate and multivariate logistic regression analysis were used to analyze the influencing factors of PE after hepatectomy. The receiver operating characteristic (ROC) curve was used to analyze the effect of each index on predicting PE after hepatectomy. Results:Multivariate logistic regression analysis showed that red blood cell volume distribution width ( OR=1.399, 95% CI: 1.223-1.601) and systemic inflammation response index (SIRI) on postoperative day (POD) 3 ( OR=1.317, 95% CI: 1.124-1.544) and D-dimer on POD1 ( OR=1.208, 95% CI: 1.026-1.421) were associated with a higher risk of PE after hepatectomy (all P<0.05). The area under the ROC curve of SIRI on POD3, D-dimer on POD1, and red blood cell volume distribution width on POD3 on predicting PE after hepatectomy was 0.763 (95% CI: 0.689-0.838), 0.744 (95% CI: 0.668-0.820) and 0.796 (95% CI: 0.727-0.864), respectively. Conclusion:Red blood cell volume distribution width on POD3, SIRI on POD3 and D-dimer on POD1 are the risk factors of PE after hepatectomy. These three indicators have certain predictive value for PE after hepatectomy.

Objective:To investigate the clinical characteristics and independent risk factors of severe disease in patients with anti-nuclear matrix protein (NXP) 2 antibody-positive juvenile dermatomyositis (JDM).Methods:A retrospective cohort study was conducted, including 219 anti-NXP2 antibody-positive JDM patients admitted to 23 children′s hospitals across China from July 2011 to July 2023. Patients were classified into severe and non-severe groups based on classification criteria for severe dermatomyositis. Demographic characteristics, clinical manifestations, and laboratory parameters were compared between the 2 groups using independent sample t-test, Mann-Whitney U test, or χ2 test. Univariate and multivariate Logistic regression analyses were performed to identify risk factors for severe disease. The receiver operating characteristic curve was employed to calculate optimal cut-off values. Results:Among the 219 patients, 108 were male and 111 were female, with an age at onset of 6.3 (3.5, 9.4) years. The severe group comprised 69 patients, and the non-severe group 150 patients. The severe group had significantly higher rates of fever, heliotrope rash, subcutaneous edema, periorbital edema, anti-Ro52 antibody positivity, as well as elevated levels of ferritin-to-albumin ratio (FAR), creatine kinase (CK), aspartate aminotransferase (AST), and lactate dehydrogenase (LDH) (all P<0.05). Multivariate analysis identified anti-Ro52 antibody positivity ( OR=13.26, 95% CI 1.37-128.29) and elevated FAR ( OR=1.90, 95% CI 1.09-2.31) as independent risk factors for severe anti-NXP2 antibody-positive JDM (both P<0.05). Receiver operating characteristic curve analysis revealed that a FAR cutoff value of 6.82 predicted severe disease with an area under the curve of 0.87 (95% CI 0.81-0.94, P<0.001), sensitivity of 0.85, and specificity of 0.70. All patients received glucocorticoid therapy, and the severe group received higher proportions of steroid pulse therapy, cyclophosphamide, mycophenolate mofetil, intravenous immunoglobulin, biologics, and adjuvant treatments compared to the non-severe group (all P<0.05). In terms of outcomes, 2 patients (2.9%) in the severe group died (due to neurological involvement and intestinal perforation, respectively), while the remaining patients achieved complete clinical response or remission. All patients in the non-severe group achieved remission. Conclusions:The primary clinical features of anti-NXP2 antibody-positive JDM included fever, heliotrope rash, subcutaneous edema, periorbital edema, anti-Ro52 antibody positivity, and elevated levels of CK, AST, LDH, and FAR. Furthermore, anti-Ro52 antibody positivity and a FAR>6.82 were identified as independent risk factors.

Objective:To investigate the clinical and genetic variation characteristics of a child with autosomal dominant lateral temporal lobe epilepsy caused by de novo variation of the MICAL1 gene. Methods:Clinical data of the patient with autosomal dominant lateral temporal lobe epilepsy caused by MICAL1 gene variation diagnosed in Children′s Hospital of Soochow University in August 2019 were collected. The whole exome sequencing was performed on the core members of the family, and the characteristics of gene variations were analyzed. Results:The proband, a 10 years and 5 months old boy, was admitted to the hospital because of "intermittent convulsions for 7 years". The clinical manifestations included focal or generalized tonic-clonic seizures and hearing aura, with normal language and intellectual development. No abnormalities were found in the T 1 and fluid attenuated inversion recovery sequences of the cranial 3.0 T magnetic resonance imaging and 3D thin-slice magnetic resonance imaging.Long-range video electroencephalogram showed the distribution of spinous and slow spinous waves in the left frontal and temporal areas. The results of whole exome gene sequencing in the core family members showed heterozygous de novo missense variation in the MICAL1 gene of the proband (NM_022765): c.763G>T(exon6)(p.Val255Leu) that had not been reported. According to American College of Medical Genetics and Genomics and Association for Molecular Pathology guidelines (2015), the mutation was considered potentially pathogenic. The application of antiepileptic drugs was effective in controlling epileptic seizures. Conclusions:Auditory symptoms are main clinical manifestations for the child with autosomal dominant lateral temporal lobe epilepsy. Antiepileptic drugs can effectively control epileptic seizures of the child, and the MICAL1 gene c.763G>T (p.Val255Leu) mutation is the genetic cause of the proband.

Objective:To summarize the clinical characteristics of 5 children with developmental epileptic encephalopathy type 17 (DEE17) caused by GNAO1 gene variants confirmed by whole-exome sequencing and analyze the features of their genetic variants. Methods:A retrospective analysis was conducted on the clinical data of 5 children diagnosed with GNAO1-related DEE17 in the Department of Neurology, Children′s Hospital of Soochow University from January 2019 to October 2024. Their clinical features, genetic testing results, neuroimaging findings, electroencephalogram (EEG) results, and treatment regimens were summarized. Follow-up was performed via telephone or outpatient visits. Results:Among the 5 diagnosed children (3 males, 2 females), the age of onset ranged from 2 days to 2 years, and the age at diagnosis ranged from 2 days to 6 years. Four children presented with seizures in the neonatal or infantile period, manifesting as hypotonia, developmental delay, and seizure types including generalized tonic-clonic, myoclonic, and epileptic spasms. One child had a later onset at 2 years, presenting with language delay, intellectual disability, and involuntary movements, followed by seizures at 6 years, including focal and generalized tonic-clonic seizures. Genetic testing revealed de novo heterozygous missense variants in GNAO1 in all 5 cases: c.119G>C (p.G40A), c.808A>C (p.N270H), c.808A>G (p.N270D), c.118G>C (p.G40R), and c.17G>T (p.S6I). Among these variants, c.119G>C and c.17G>T were previously unreported pathogenic variants. Neuroimaging showed nonspecific changes in 3 children (widened frontal-temporal subarachnoid space, delayed myelination) and abnormal white matter signals in 2 cases. Long-term video-EEG revealed abnormal discharges and background slowing in all cases: multifocal discharges in 4 cases and focal epileptiform discharges (left mid-temporal) in 1 case. Clinical seizures were captured in 3 cases: 1 with a burst-suppression pattern and 2 with hypsarrhythmia. All patients received 3 or more antiseizure medications. Four cases (cases 1-4) responded well to topiramate combination therapy, with 2 cases (cases 1, 2) achieving complete seizure freedom and 2 cases (cases 3, 4) experiencing more than a 50% reduction in seizures. One child (case 3) achieved seizure control with an adjunctive ketogenic diet. The late-onset case (case 5) required a combination of levetiracetam, oxcarbazepine, and valproate for seizure management. Conclusions:GNAO1 variants can lead to DEE17 with diverse seizure types, often requiring multiple antiseizure medications, among which topiramate is effective. Early-onset cases typically present with seizures and developmental delay, while late-onset cases may exhibit language delay, intellectual disability, movement disorders, and refractory epilepsy. Genetic testing should be performed early for timely diagnosis.

Objective To construct a prognostic model of future asthma exacerbation risk in adults by com-bining novel biomarkers of serum chitinase-3-like protein 1(YKL-40),dipeptidyl peptidase-4(DPP4)and conventional predictors.Methods Patients with asthma in the non-acute exacerbation phase were recruited from the People's Hospital of Yubei District of Chongqing,from March 2022 to May 2023.Baseline clinical da-ta collected included medical history,forced expiratory volume in the first second(FEV1)/forced vital capacity(FVC),percentage of predicted forced expiratory volume in the first second(FEV1％pred),blood eosinophil count(EOS),blood neutrophil count(NEU),fractional exhaled nitric oxide(FeNO),serum YKL-40,and ser-um DPP4,etc.The patients were followed for one year to gather data on asthma acute exacerbations and their timings as defined in this study.A COX proportional hazards regression model was used to construct a prog-nostic model for future asthma exacerbations,with internal validation and results presentation.Results A to-tal of 224 patients with asthma completed the study.During the one-year follow-up period,102 patients experi-enced acute exacerbations as defined in this study.Based on univariate COX regression,stepwise regression for variable selection,clinical significance,and model simplicity,asthma control test(ACT)score group,number of asthma exacerbations in the past year group,log10(YKL-40),log10(FeNO),log10(EOS),and FEV1％pred were the following predictors were included in the final model.The overall C-statistic of the model was 0.795(95％CI:0.754-0.836),the area under the curve at the 52-week follow-up was 0.879(95％CI:0.834-0.924),and the Brier score at the 52-week follow-up was 0.142(95％CI:0.117-0.168).The calibration curve was close to a slope of 1,and bootstrap validation suggested good stability of the prediction model.The model was presented using a Nomogram and a dynamic scoring table in a web APP,which can be used to predict the risk of asthma exacerbations within 52 weeks for individual patients.Conclusion The prediction model based on serum YKL-40,EOS,FeNO,the number of asthma exacerbation in the past year group,FEV1％pred and ACT scores group can accurately predict the probability of acute attacks in 52 weeks of asthma patients.

Objective To investigate the clinical characteristics,imaging features,laboratory test results,and prognosis of children with acute lymphoblast leukemia(ALL)complicated by cerebral hemorrhage.Methods A retrospective analysis was conducted on the clinical data of 20 children with ALL complicated by cerebral hemor-rhage admitted to the Department of Hematology,Children's Hospital of Soochow University from June 20,2014 to June 20,2024.Results The clinical manifestation of the 20 children with ALL complicated by cerebral hemorrhage were complex and diverse,with disturbance of consciousness being the most common initial symptom.The prognosis varied depending on the size and location of the hematoma and whether it ruptured into the ventricle.Among the 20 cases,14(70％)demonstrated improvement in intracranial lesions,with 8(40％)cases exhibiting substantial lesion absorption and favorable prognosis.Six cases(30％)showed improvement in intracranial lesions but not complete resolution,three cases developed focal encephalomalacia,two cases had residual symptomatic epi-lepsy and one had residual right-sided hemiplegia.Furthermore,three(15％)cases suffered recurrent cerebral hemorrhages at distinct locations from the initial event following improvement of the primary hemorrhage,and 3(15％)cases led to mortality.Conclusions Neurological symptoms in children with acute lymphoblast leukemia(ALL)complicated by cerebral hemorrhage are diverse and often atypical.Timely cranial imaging and laboratory tests are necessary,while surgical intervention and platelet transfusion should be a prudential consideration.

Objective To propose a classification model based on a pseudo-bag strategy and feature adjustment for whole slide imaging in pathology.Methods A pseudo-bag generator was constructed to divide a parent bag into 3 pseudo-bags for increasing the number of training bags.Then,a pseudo-bag learning method based on Nystr?m-based algorithm for approximating self-attention and a selective feature fusion method were employed to process the pseudo-bags.Specifically,the pseudo-bag learning method based on Nystr?m-based algorithm for approximating self-attention reduced computational complexity through an improved multi-head self-attention mechanism while deeply extracting instance features to obtain pseudo-bag classification predictions,thereby enhancing pseudo-bag classification accuracy;and the selective feature fusion method refined pseudo-bag features by filtering and extracting relevant instances.Finally,the model adjusted bag features by extracting confounding factors to avoid interference from irrelevant information and further improve classification accuracy.Results The proposed model was evaluated on two datasets(CAMELYON-16 and TCGA-NSCLC)and compared with 10 other methods,and the results demonstrated that the proposed model achieved the best performance.The proposed method reached an accuracy of 0.943 on the CAMELYON-16 dataset and 0.906 on the TCGA-NSCLC dataset.Conclusion The proposed model can significantly improve the accuracy of whole-slide pathological image classification by effectively mitigating the overfitting and avoiding interference from irrelevant information.

A VLi-net based cell nucleus segmentation method integrating convolutional neural networks(CNN)and Vision Transformer(ViT)is proposed to address the limitation that the U-Net with CNN as its backbone is only proficient in capturing local features and has a restricted receptive field.Firstly,to mitigate challenges such as high cost of data annotation and insufficient annotated data,text annotations are introduced to enhance the network's understanding of image information.Secondly,to improve the segmentation performance of VLi-net,ViT and CNN are combined to fully extract global and local features,with multi-receptive field convolution features incorporating into the ViT structure for effectively mitigating the issues of limited local information interaction and single feature representation in ViT.Finally,an interactive fusion module(ViFusion)is used to efficiently fuse the multi-level features from the CNN and ViT branches.Experimental results show that VLi-net achieves a Dice coefficient of 80.85%and a mean intersection over union(MIoU)of 66.83%on the MoNuSeg dataset,obtains a Dice coefficient of 80.53%and a MIoU of 67.54%on the DSB-2018 dataset,and has a Dice coefficient of 86.87%and a MIoU of 77.44%on the TNBC dataset.These findings confirm that VLi-net outperforms other methods across multiple experimental metrics.

A VLi-net based cell nucleus segmentation method integrating convolutional neural networks(CNN)and Vision Transformer(ViT)is proposed to address the limitation that the U-Net with CNN as its backbone is only proficient in capturing local features and has a restricted receptive field.Firstly,to mitigate challenges such as high cost of data annotation and insufficient annotated data,text annotations are introduced to enhance the network's understanding of image information.Secondly,to improve the segmentation performance of VLi-net,ViT and CNN are combined to fully extract global and local features,with multi-receptive field convolution features incorporating into the ViT structure for effectively mitigating the issues of limited local information interaction and single feature representation in ViT.Finally,an interactive fusion module(ViFusion)is used to efficiently fuse the multi-level features from the CNN and ViT branches.Experimental results show that VLi-net achieves a Dice coefficient of 80.85%and a mean intersection over union(MIoU)of 66.83%on the MoNuSeg dataset,obtains a Dice coefficient of 80.53%and a MIoU of 67.54%on the DSB-2018 dataset,and has a Dice coefficient of 86.87%and a MIoU of 77.44%on the TNBC dataset.These findings confirm that VLi-net outperforms other methods across multiple experimental metrics.

Objective To propose a classification model based on a pseudo-bag strategy and feature adjustment for whole slide imaging in pathology.Methods A pseudo-bag generator was constructed to divide a parent bag into 3 pseudo-bags for increasing the number of training bags.Then,a pseudo-bag learning method based on Nystr?m-based algorithm for approximating self-attention and a selective feature fusion method were employed to process the pseudo-bags.Specifically,the pseudo-bag learning method based on Nystr?m-based algorithm for approximating self-attention reduced computational complexity through an improved multi-head self-attention mechanism while deeply extracting instance features to obtain pseudo-bag classification predictions,thereby enhancing pseudo-bag classification accuracy;and the selective feature fusion method refined pseudo-bag features by filtering and extracting relevant instances.Finally,the model adjusted bag features by extracting confounding factors to avoid interference from irrelevant information and further improve classification accuracy.Results The proposed model was evaluated on two datasets(CAMELYON-16 and TCGA-NSCLC)and compared with 10 other methods,and the results demonstrated that the proposed model achieved the best performance.The proposed method reached an accuracy of 0.943 on the CAMELYON-16 dataset and 0.906 on the TCGA-NSCLC dataset.Conclusion The proposed model can significantly improve the accuracy of whole-slide pathological image classification by effectively mitigating the overfitting and avoiding interference from irrelevant information.