Objective:To establish three machine learning prediction models based on CT imaging characteristics of papillary thyroid carcinoma (PTC) , and use SHAP (shapley additive explanations) analysis to investigate the contribution of each CT image features in the best model.Methods:CT imaging features in 426 cases of 440 PTCs confirmed pathologically from Jan. 2016 to Jan. 2021 at the affiliated Hangzhou First People’s Hospital of Westlake University Medical School were retrospectively analyzed. compared with 467 cases of 528 nodular goiter (NG) , evaluating the distribution of four CT characteristics: cookie bite sign, enhanced range of narrowing/blur (ERNB) , microcalcifications, and irregular shape. We split the data into 8∶2 ratio for training and testing sets, then constructed three machine learning models using XGBoost, RF, and SVM. Based on AUC, accuracy, F1 score, and other metrics, we selected the best model. Lastly, we used SHAP values to assess each CT feature’s contribution and positive/negative effects on the model.Results:Among 440 PTC and 528 NG nodules, CT features like cookie bite sign, ERNB, microcalcifications, and irregular shape occurred in 326 and 30 ( χ 2=483.05, P<0.001) , 363 and 106 ( χ 2=374.45, P<0.001) , 158 and 53 ( χ 2=94.24, P<0.001) , and 354 and 52 ( χ 2=491.34, P<0.001) nodules, respectively. The machine learning models built using XGBoost, RF, and SVM had AUC, accuracy, and F1 scores ranging from 0.884~0.925, 0.867~0.873, and 0.844~0.854 respectively on the training set. On the test set, the scores ranged from 0.869~0.923, 0.845~0.871, and 0.803~0.845. Among them, the XGBoost model demonstrated the highest diagnostic performance on the test set. Among the four CT features, irregular shape had the highest absolute SHAP value, positively contributing to PTC diagnosis. Conclusion:XGBoost model showed the highest PTC diagnostic performance. Irregular shape had the greatest positive impact on PTC diagnosis.

Objective:To investigate the changes of lymphocyte to monocyte ratio (LMR), matrix metalloproteinase-2 (MMP-2), hypoxia inducible factor-1 (HIF-1 α) in patients with breast cancer undergoing surgical resection before and after radiotherapy and their relationship with short-term prognosis.Methods:106 patients with breast cancer who underwent surgical resection and radiotherapy from Jan. 2021 to Jan. 2024 were analyzed as the study subjects. All patients received the same scheme of radiotherapy. The changes of LMR, MMP-2, and HIF-1 α levels at different time points before and after radiotherapy were recorded, and the clinical efficacy of the two groups of patients after radiotherapy was compared. According to the short-term efficacy after radiotherapy, the patients were divided into 31 patients with poor efficacy (progression+deterioration) and 75 patients with good efficacy (complete remission+effective remission+partial remission). The basic clinical data of the two groups were compared. The binary regression analysis was used to analyze the factors that affect the short-term prognosis of patients with breast cancer who had undergone surgery after radiotherapy. A risk prediction model was constructed, and the ROC analysis model was used to predict the value.Result:Compared with before radiotherapy, LMR decreased and MMP-2 and HIF-1 α increased after one week of radiotherapy ( t=2.68, -2.76, -1.96, P=0.008, 0.006, 0.052); After one course of radiotherapy and three courses of radiotherapy, LMR increased and MMP-2 and HIF-1 α decreased ( t=-3.02, 5.14, 5.86, all P<0.05; t=-7.95, 19.80, 21.36, all P<0.001). Low expression of LMR, high expression of MMP-2 and high expression of HIF-1 α are independent risk factors for poor short-term prognosis of breast cancer patients after radiotherapy. LMR、MMP-2、HIF-1α、The prediction model can predict the poor short-term prognosis of breast cancer patients after radiotherapy. The AUC values were 0.780 (95% CI: 0.674~0.887), 0.759 (95% CI: 0.659~0.859), 0.840 (95% CI: 0.748~0.931), and 0.887 (95% CI: 0.808~0.965), respectively. When taking the cut-off values, the sensitivities were 0.853, 0.903, 0.677, and 0.920, and the specificities were 0.677, 0.533, 0.933, and 0.774, respectively. Bootstrap method (B=1000) was used to perform internal validation on the prediction model of poor short-term prognosis of breast cancer patients after radiotherapy. The results showed that the prediction curve after deviation correction was close to the ideal state, and the C-index reached 0.774, indicating that the model had strong prediction ability. In addition, the decision curve of the model shows that the net profit is always positive and better than the two invalid reference lines within the threshold probability range of 0.1 to 1.0. Conclusion:LMR of postoperative radiotherapy patients with breast cancer decreased first and then increased, MMP-2 and HIF-1 α increased first and then decreased, all of which are independent predictors of short-term prognosis. The combined model has a significant predictive effect on poor efficacy and can be used as an important reference for clinical prognosis evaluation.

Objective To observe the value of non-contrast CT radiomics extreme gradient boosting(XGBoost)model based on SHAP method for predicting acute necrotic collection(ANC)around acute pancreatitis(AP).Methods A total of 307 patients with initially clinically diagnosed AP were retrospectively enrolled.The optimal radiomics features of peripheral pancreatic tissue volume of interest(VOI)were extracted and screened based on automatic segmentation on the first non-contrast CT,and the evaluation results of modified CT severity index(MCTSI)score of AP severity based on first enhanced CT were recorded.The patients were divided into peripancreatic ANC group(ANC group)and acute peripancreatic fluid collection(APFC)group according to follow-up abdominal CT.XGBoost method was used to construct radiomics model,MCTSI model and combined model for predicting AP ANC based on the optimal radiomics features,MCTSI and their combination,respectively.The diagnostic efficacy of each model was evaluated using 5-fold cross-validation method,and the contribution of each variable to combined model was analyzed with SHAP method.Results Among 307 cases,there were 134 cases in ANC group and 173 in APFC group.Totally 6 optimal radiomics features were screened based on the first non-contrast CT.The area under the receiver operating characteristic curve(AUC)of radiomics model,MCTSI model and combined model was 0.936,0.693 and 0.917,respectively.The AUC of MCTSI model was lower than that of radiomics model and combined model(Z=-3.485,-2.824,both P＜0.01),while no significant difference of AUC was found between radiomics model and combined model(Z=-0.817,P=0.415).The contribution of optimal radiomics features to combined model were all higher than that of MCTSI score.Conclusion Non-contrast CT radiomics XGBoost model could effectively predict AP ANC.

Objective To study the three-dimensional changes of soft and hard tissue in male patients with unilateral non-syndromic cleft lip and palate in the mixed dentition period before and after maxillary protraction.Methods Twenty male patients with unilateral non-syndromic cleft lip and palate in the mixed dentition period treated by maxillary anterior traction in the Department of Orthodontics of Affiliated Stomatological Hospital of Nanjing Medical University were selected(average age(10.6±1.23)years old).Cone beam CT was taken before and after treatment.Dolphin 3D 11.95 software was used for three-dimensional measurement and analysis.SPSS 25.0 software package was used for statistical analysis.The self-controlled paired t test was used to compare the changes in soft and hard tis-sues of male patients with unilateral cleft lip and palate before and after treatment.The changes in the anterior displacement of the ANS point,the anterior displacement of point A,and the posterior displacement of point B were tested using the one-sample t test.Results The sagittal skeletal changes were significantly increased in ∠ SNA(P＜0.01),∠ANB(P＜0.01),Y axis(P＜0.05),the forward displacement of ANS point(P＜0.01)and A point(P＜0.01)and the backward displacement of B point(P＜0.01),but ∠SNB(P＜0.05)was decreased significantly.The vertical skeletal changes showed that ∠MP-FH(P＜0.01),∠MP-SN(P＜0.05)and the dis-tance of ANS-Me(P＜0.05)were increased significantly,but ∠SN-PP(P＜0.01)was decreased significantly.The dental changes inclu-ding ∠U1-NA(P＜0.01),the distance of U1-NA(P＜0.01),∠U1-SN(P＜0.01),overjet and the Wits were increased significantly,but ∠ L1-NB(P＜0.01),the distance of L1-NB(P＜0.01)and ∠L1-MP(P＜0.01)were decreased significantly.The changes of soft tis-sue including ∠ S-Ns-Sn(P＜0.01),∠ Sn-Ns-Bs(P＜0.01),the distance of UL-EP(P＜0.01)and LL-UL(P＜0.01)were increased sig-nificantly.Conclusion After the treatment of maxillary protraction,the forward growth of maxilla will be possibly promoted on patients with cleft lip and palate in the peak of growth timing,as well as the intermaxillary relationship and soft tissue profile,but the side effects should be paid attention to.

Objective:To observe effectiveness of combining diffusion tensor imaging (DTI) with motor evoked potentials (MEPs) in evaluating the upper limb motor functioning of stroke survivors.Methods:Thirty-seven stroke survivors with upper limb motor dysfunction were selected. At the 4th, 12th and 24th week after their onset, each was were assessed using Fugl-Meyer Upper Limb (FMA-UE) scoring, the National Institutes of Health stroke scale (NIHSS), the modified Rankin Scale (mRS), the Barthel Index (BI) and hemiplegic hand function classification. DTI was also applied and MEPs were measured. The patients were divided into an MEP positive group and an MEP negative group according to the existence of the MEP waveform. The DTI and MEP parameters were correlated with the FMA-UE scores, linear regressions were evaluated and a receiver operating characteristics curve was prepared to estimate the utility of DTI in predicting hand function. The sensitivity and specificity of MEPs in predicting hand function recovery were evaluated.Results:The asymmetry index (FAa) of the average anisotropy score of the posterior limb of the internal capsule and the FAa of the cerebral peduncle were both significantly correlated with the FMA-UE scores at the 12th and 24th weeks. The best cut-off points for predicting functional recovery of a patient′s hand were 0.155 for the FAa of the posterior limb of the internal capsule and 0.145 for the cerebral peduncle. Among the 37 patients, the MEPs of 8 (the MEP positive group) could be extracted, and their hand functions recovered completely. The sensitivity of the MEPs in predicting the complete recovery of hand function was 80% with 100% specificity. The linear regression analysis showed 77% prediction accuracy for the FAa and MEPs of the cerebral peduncle for upper limb motor function at the 24th week after onset. In the MEP negative group, two patients completely recovered their hand function, with one′s FAa less than 0.145, and the other′s more than 0.145. When the MEP was negative, the sensitivity of DTI in predicting the recovery of hand function was 50% with 81.5% specificity.Conclusions:DTI combined with MEPs can be used as an index to evaluate the prognosis of upper limb motor function in stroke patients.

This study was aimed at investigating the effectiveness of various host nucleic acid removal and non-specific amplifica-tion techniques in animal tissue samples,to increase the accuracy of pathogen identification in tissue samples.Simulated samples were prepared with a mixture of mouse lung tissue homogenates and Klebsiella pneumoniae fluids,and processed with six host nucleic acid removal kits and three non-specific amplification techniques.The effectiveness of each method in removing host DNA and enriching nucleic acids of pathogenic microorganisms was evaluated through real-time fluorescence quantitative PCR and high-throughput se-quencing.For host nucleic acid removal techniques,the method of selective cleavage and quantitative degradation of host DNA(Com-plete5 kit)effectively decreased the host nucleic acid content in tissue samples and increased the relative abundance of pathogen nucleic acids.In contrast,the magnetic bead method for host DNA removal(Next microbiome DNA enrichment Kit kit)was less effec-tive.At lower pathogen concentrations(77 CFU/mL),the Vazyme kit was more effective than the other kits in removing host nucleic acids.Non-specific amplification techniques(MALBAC whole genome amplification,MDA isothermal amplification,and random primer amplification)were not applicable to tissue samples and were not effective in increasing the relative abundance of pathogen nucleic acids.Selective lysis and quantitative degradation of host DNA were suitable for processing tissue samples with high host back-ground and low pathogenic microorganism levels,whereas non-specific amplification methods were not applicable to tissue samples for pre-processing of macro-genome high-throughput sequencing.

Objective To address the personalized differences in motion gait between stroke patients and healthy older adults,as well as the issue of abnormal gait recognition,a deep learning fusion-based approach is proposed to effectively improve the accuracy of abnormal gait recognition.Methods A model fusing convolutional neural networks(CNN)and bidirectional long short-term memory networks(BiLSTM)was adopted,with the introduction of a residual network(ResNet).Unilateral ankle joint movement data at different walking speeds within a comfortable range were collected from healthy older adults and stroke patients.Signals from inertial sensors and electromyography sensors were used as inputs,while gait features were analyzed and gait differences between the two groups were compared.The effectiveness of the model was validated by comparing the classification performance of traditional deep learning models and CNN-ResNet-BiLSTM models with different layer combinations in terms of abnormal gait recognition accuracy.Results The CNN-ResNet-BiLSTM model,which introduced residual connectivity,performed excellently in abnormal gait recognition.Compared with traditional deep learning models such as the gated recurrent unit(GRU)and long short-term memory network(LSTM),its prediction accuracy was improved by 13.6％and 8.36％,respectively.Additionally,compared with other model combinations,this model achieved an overall accuracy of 97.78％.Conclusions The algorithm proposed in this study can be applied to stroke-related abnormal gait detection,providing technique support for the early diagnosis and precise monitoring of such diseases.

OBJECTIVE:The quality of systematic reviews/meta-analyses directly affects the reliability of clinical decision-making basis.Currently,there is no literature quality research on the systematic reviews/meta-analyses of platelet rich plasma versus hyaluronic acid in the treatment of knee osteoarthritis both domestically and internationally.This article will comprehensively evaluate the methodological quality,reporting quality,and evidence quality of the systematic reviews/meta-analyses of platelet rich plasma versus hyaluronic acid in the treatment of knee osteoarthritis.METHODS:Computer searches were conducted on CNKI,WanFang Data,VIP,CBM,PubMed,Embase,and The Cochrane Library.From the establishment of the database until January 24,2024,all systematic reviews/meta-analyses treated with platelet rich plasma versus hyaluronic acid for knee osteoarthritis were collected.Two evaluators independently conducted literature screening and data extraction,and used the AMSTAR 2,PRISMA 2020,and GRADE systems to evaluate and summarize the methodological,reporting,and evidence quality of the included systematic reviews/meta-analyses.RESULTS:A total of 18 qualified systematic reviews/meta-analyses were included,and the results showed that the efficacy and safety of platelet rich plasma might be better than that of hyaluronic acid.The methodological quality of all 18 studies was extremely low;4 reports had poor quality and relatively serious information defects,while 14 reports had moderate quality and some information defects.Among the 275 outcome measures of 16 systematic reviews/meta-analyses,there were 9 medium quality evidence,90 low-quality evidence,and 176 extremely low-quality evidence,with no high-quality evidence.CONCLUSION:At present,the quality of systematic reviews/meta-analyses literature on the treatment of knee osteoarthritis with platelet rich plasma versus hyaluronic acid is relatively low.In the future,the authors of the systematic reviews need to strictly follow the entries of quality evaluation tools such as AMSTAR 2 and PRISMA 2020 in terms of plan registration,research type explanation,retrieval strategy,exclusion list,research site and funding source,bias risk analysis,publication bias evaluation,and public information acquisition,and conduct evidence quality evaluation on the combined results of the systematic reviews/meta-analyses to provide more reliable and rigorous evidence-based basis for clinical practice.

Objective:To explore the therapeutic effect and mechanism of Baicalin(BG)on chronic migraine(CM)rats.Methods:CM rat model was established by intermittent subcutaneous injection of nitroglycerin.Rats were randomly grouped into con-trol group,chronic migraine group(CM group),low-dose and high-dose Baicalin groups(BG-L group,BG-H group),Baicalin+MIP-1α recombinant protein group(BG+MIP-1α group)and CCR1 inhibitor ZK811752 group(CCR1-ZK group),with 10 rats in each group.General behavior of rats were observed and scored;von Frey fiber optic pain gauge was applied to measure the mechanical pain threshold of rats;Nissl staining was applied to observe the neuronal structure of ventrolateral periaqueductal gray(vlPAG);ELISA was applied to detect levels of TNF-α and IL-1β;immunohistochemical method was applied to detect expressions of pain related pro-teins c-Fos and CGRP;Western blot was applied to detect expressions of MIP-1α and CCR1 proteins.Results:Compared with Control group,the body mass,mechanical pain threshold,and the number of Nissl bodies in vlPAG area of rats in CM group were obviously reduced,the general behavioral score,levels of TNF-α,IL-1β,and expressions of c-Fos,CGRP,MIP-1α and CCR1 were obviously increased(P<0.05);compared with CM group,the body mass,mechanical pain threshold,the number of Nissl bodies in vlPAG area of rats in BG-L group,BG-H group and CCR1-ZK group were obviously increased,the general behavioral score,levels of TNF-α,IL-1β,and expressions of c-Fos,CGRP,MIP-1α and CCR1 were obviously reduced(P<0.05);compared with BG-H group,the body mass,mechanical pain threshold,and the number of Nissl bodies in vlPAG area of rats in BG+MIP-1α group were obviously reduced,general behavioral score,levels of TNF-α,IL-1β,and expressions of c-Fos,CGRP,MIP-1α and CCR1 were obviously increased(P<0.05).Conclusion:Baicalin has certain therapeutic effect on chronic migraine rats,and its mechanism may be related to inhibiting the activation of MIP-1α/CCR1 signaling pathway.

Inflammatory bowel disease (IBD) is a chronic inflammatory disease that occurs in the gastrointestinal tract and adversely affects the quality of life of patients and the healthcare burden on society. Currently, drug therapy alone can achieve clinical remission in some patients, but it cannot break through the ceiling of what can be achieved with current therapies, and new therapeutic options still need to be found. Neuro-immunity-inflammation is closely related to each other, and the potential role of neuroelectric stimulation in modulating inflammatory response has made it a hot spot in the study of IBD treatment. However, the intrinsic mechanism of action is still unexplained and needs to be explored by further studies. This review discusses the role and mechanisms of vagus nerve electrical stimulation, sacral nerve electrical stimulation, tibial nerve electrical stimulation and other neuromodulation methods.