ObjectiveTo investigate the effect of different exercise interventions on metabolism and liver parameters in patients with metabolic dysfunction-associated fatty liver disease （MAFLD）， and to provide evidence-based recommendations for clinical exercise rehabilitation. MethodsThis study was conducted according to the PRISMA guidelines， and the protocol was registered on the PROSPERO platform， with a registration number of CRD42025641717. PubMed， Web of Science， Scopus， Wiley Online Library， CNKI， Wanfang Data， and VIP were searched for related articles published up to September 2024. The Cochrane tool for assessing risk of bias was used to assess the quality of articles， and Stata MP 17.0 was used to perform the network meta-analysis. ResultsA total of 57 articles were included， involving 2 648 patients. The results showed that aerobic exercise combined with resistance exercise had the best effect in improving body mass index （mean difference ［WMD］=-0.97， 95% confidence interval ［CI］： -1.66 to -0.28］， P<0.05， surface under the cumulative ranking curve ［SUCRA］=85.4） and triglycerides （WMD=-29.6， 95%CI： -46.66 to 12.54， P<0.05， SUCRA=87.3）； resistance exercise was the optimal intervention method for improving total cholesterol （WMD=-15.99， 95%CI： -24.19 to -7.79， P<0.05， SUCRA=79.9） and glutamine transaminase （WMD=-8.08， 95%CI： -12.13 to -4.02， P<0.05， SUCRA=87.3）； low-intensity aerobic exercise had the best effect in improving aspartate aminotransferase （WMD=-4.3， 95%CI： -8.45 to -0.15， P<0.05， SUCRA=73.5）， gamma-glutamyl transpeptidase （GGT） （WMD=-3.26， 95%CI： -7.79 to 1.27， P>0.05， SUCRA=82.3）， and glycated hemoglobin （HbA1c） （WMD=-0.6， 95%CI： -2.02 to 0.82， P>0.05， SUCRA=78.8）； moderate-intensity aerobic exercise was the optimal intervention modality to improve Homeostasis Model Assessment of Insulin Resistance （WMD=-0.92， 95%CI： -1.51 to -0.33， P<0.05， SUCRA=69.4）. It should be noted that there were no significant differences in HbA1c and GGT across different exercise interventions （all P>0.05）， suggesting that there was currently no sufficient statistical evidence to support that exercise could improve these two indicators. ConclusionBased on the comprehensive league table and cumulative probability ranking， aerobic exercise combined with resistance exercise， resistance exercise， and low- and moderate-intensity aerobic exercise may be the best exercise modality for improving key indicators in MAFLD patients， and targeted exercise modalities should be selected for intervention against different indicators； however， due to limitations of the original studies， further studies are needed for validation and exploration.

Objective To investigate the expansion margins of the planning target volume (PTV) and the planning organ at risk volume (PRV) in nasopharyngeal carcinoma patients immobilized with Styrofoam and head-neck-shoulder mask. Methods A convenient sample of 33 nasopharyngeal carcinoma patients who received radiotherapy at Huanggang Central Hospital from January to October 2024 were selected as the research subjects. All patients underwent cone beam CT scans during the first three treatments and weekly thereafter. After registration and calibration, the setup errors in the X (LAT), Y (LNG), and Z (VRT) directions were recorded. Statistical analysis was performed on the setup errors in each direction to determine differences, and the expansion margins for PTV and PRV were calculated using empirical formulas. Results A total of 229 cone beam CT images were collected. Statistical analysis found that the setup errors (systematic error ± random error) of the patients in the X, Y, and Z directions were 1.05 ± 0.72, 1.30 ± 0.80, and 1.29 ± 0.82 mm, respectively. The expansion margins for PTV in the left-right, superior-inferior, and anterior-posterior directions were 1.40, 1.76, and 1.8 mm, respectively. The expansion margins for PRV in these directions were 0.83, 1.02, and 1.05 mm, respectively. Conclusion For patients immobilized using Styrofoam and head-neck-shoulder mask, it is recommended that the expansion margins for PTV and PRV be set at 2 mm and 1 mm, respectively, in the left-right, superior-inferior, and anterior-posterior directions, and the PRV margin for the spinal cord be set at 3 mm in all directions.

Objective:To compare the clinical and radiological outcomes between the ankle dislocation exposure versus the traditional posterolateral approach exposure in the treatment of complex posterior ankle malunion.Methods:This retrospective study analyzed the 50 patients (50 feet) with posterior ankle malunion who had been treated at Department of Foot and Ankle Orthopedics, Foshan Hospital of Traditional Chinese Medicine Affiliated to Guangzhou University of Chinese Medicine from May 2015 to June 2023. This cohort included 18 males and 32 females, with an age of (46.2±12.0) years. The duration from injury to surgery averaged 7.0 (4.0, 12.2) months. The patients were divided into 2 groups based on the surgical exposure methods: an ankle dislocation group (22 cases) in which the posterior ankle malunion was exposed by ankle dislocation before reduction and fixation, and a traditional posterolateral approach group (28 cases) in which the posterior ankle malunion was exposed by the traditional posterolateral approach before reduction and fixation. Evaluation indices included operative time, ankle arthritis grading, incidence of complications, as well as pre-operative and last follow-up recordings of distal tibial articular angle and joint step-off, visual analogue scale (VAS) pain score, and ankle-hindfoot score of American Orthopaedic Foot and Ankle society (AOFAS).Results:There was no significant difference in the baseline data between the 2 groups, indicating comparability ( P>0.05). The average follow-up was (39.4±17.9) months. The operative time in the dislocation group [(95.8±9.9) min] was significantly shorter than that in the posterolateral approach group [(121.9±14.3) min] ( P<0.05). The step-off on the distal tibial articular surface at the last follow-up in the dislocation group [0.3 (0.0, 0.9) mm] was significantly lower than that in the posterolateral approach group [1.0 (0.3, 1.1) mm] ( P<0.05). The grading of ankle arthritis was significantly worse in the posterolateral approach group than that in the dislocation group ( P<0.05). There were no differences in distal tibial articular angle, VAS pain score, AOFAS ankle-hindfoot score, or postoperative complications between the 2 groups at the last follow-up ( P>0.05). Conclusions:Both ankle dislocation exposure and traditional posterolateral approach exposure yield satisfactory clinical and radiological outcomes in patients with complex posterior ankle malunion. However, the ankle dislocation exposure offers a novel method for direct visualization of the malunion on multiple planes, which is more advantageous for restoration of the anatomical consistency of the distal tibial articular surface.

Objective To explore the application value of computed tomography(CT)based three-dimensional(3D)reconstruction technology in robot-assisted laparoscopic partial nephrectomy(RAPN)for patients with renal cell carcinoma(RCC).Methods A retrospective analysis was conducted on the clinical data of 185 RCC patients who underwent RAPN in our hospital during May 2021 and Aug.2023.Among them,98 underwent preoperative upper abdominal enhanced CT based 3D reconstruction(3D reconstruction group),and 87 did not perform it(non-3D reconstruction group).The differences in general information,tumor size,R.E.N.A.L.score,hospital stay,operation time,intraoperative blood loss,warm ischemia time,and postoperative complications were compared between the two groups.Results The 3D reconstruction group had clearer and more intuitive imaging.There were no significant differences in the general information,tumor size,R.E.N.A.L.score,intraoperative blood loss,and warm ischemia time between the two groups(P＞0.05).The 3D reconstruction group had shorter operation time[(125.8±34.9)min vs.(159.0±58.3)min],total hospital stay[(7.6±1.6)d vs.(7.7±3.0)d],and postoperative hospital stay[(3.7±0.5)d vs.(4.1±1.9)d],but longer preoperative hospital stay[(3.9±1.5)d vs.(3.5±2.2)d](all P＜0.001).Conclusion CT based 3D reconstruction technology can precisely display the anatomical parameters of tumor and its surrounding tissues,which can shorten the operation time,and postoperative and total hospital stay for RCC patients receiving RAPN.

Artificial intelligence (AI) technologies are being widely applied in radiotherapy. However, the integration of AI into clinical workflows of radiotherapy faces a series of challenges, such as poor model interpretability, domain shifts between clinical application and training data, and the inherent model uncertainties. Therefore, case-specific quality assurance (QA) is essential before deploying AI models in clinical practice. This paper reviews and summarizes QA methodologies for the application of AI models in radiotherapy across four key areas: image registration, image generation, region of interest segmentation, and treatment planning.

Objective To identify key acoustic features associated with the progression of Alzheimer's disease(AD)through voice analysis combined with machine learning and feature selection techniques,thereby constructing classification models that serve as candidate tools for the early screening of AD.Methods Voice samples from AD,mild cognitive impairment(MCI)and healthy(HC)elderly individuals were obtained from the NCMMSC2021 AD voice dataset.The voice samples underwent data preprocessing,followed by feature extraction from the eGeMAPS feature set via the OpenSmile toolkit.Classification models were obtained utilizing random forest and support vector machine(SVM)algorithms.Significance testing and feature importance ranking were conducted using Python,and the further selection of the optimal features was performed through sequential forward selection(SFS).The classification performance before and after feature selection was compared and evaluated using accuracy and the area under the receiver operating characteristic curve(AUC).Results The significant acoustic features in the classification models primarily derived from spectral slope,formant,fundamental frequency,and loudness.The optimal classification performance was achieved with the SVM model following SFS feature selection,with recognition accuracies of 0.926(AUC=0.974)for AD/MCI group,0.875(AUC=0.956)for AD/HC group,and 0.879(AUC=0.904)for MCI/HC group.Conclusion SVM model performs better than random forest model,and the use of SFS for feature selection can effectively enhance model performance.Voice analysis has the potential to serve as a valuable supplementary tool for the rapid AD assessment and screening.

Objective To analyze the effects of intravenous tirofiban following endovascular treatment on the prognosis of patients with acute anterior circulation large vessel occlusion stroke who did not achieve successful recanalization.Methods This retrospective study included consecutive patients with acute anterior circulation ischemic stroke who underwent endovascular treatment but did not achieve successful recanalization.These patients were retrospectively enrolled from the Department of Neurology at Nanjing Jinling Hospital,Affiliated Hospital of Medical School,Nanjing University,the Second Affiliated Hospital of Soochow University,Wuhan No.1 Hospital,and Yijishan Hospital of Wannan Medical College(the First Affiliated Hospital of Wannan Medical College)between January 2015 and April 2023.Baseline and clinical data were collected including age,sex,medical history(hypertension,diabetes,atrial fibrillation,hyperlipidemia),personal history(smoking and drinking),National Institutes of Health stroke scale(NIHSS)score at admission,trial of Org 10172 in acute stroke treatment(TOAST)classification(large artery atherosclerosis,cardioembolism,or other types),Alberta stroke program early CT score(ASPECTS)on admission,intravenous thrombolysis,onset-to-puncture time(OTP),collateral circulation status(poor collaterals:American Society of Interventional and Therapeutic Neuroradiology/Society of Interventional Radiology[ASTIN/SIR]score 0-2;good collateral:ASTIN/SIR score 3-4),procedure duration(from femoral or radial artery puncture to device removal),occlusion site,number of passes for thrombus retrieval,and post-procedural modified thrombolysis in cerebral infarction(mTICI)grade.Patients were divided into a tirofiban group and a non-tirofiban group based on whether they received post-procedural intravenous tirofiban.The primary outcome is a favorable functional outcome at 90-day since onset(modified Rankin scale[mRS]score ≤ 3).Safety outcomes included mortality at 90-day since onset(mRS score 6),any intracranial hemorrhage within 24 h post-procedure,and symptomatic intracranial hemorrhage within 24 h post-procedure.1∶1 propensity score matchings using the nearest neighbor method was performed based on variables with P＜0.05 in the comparison between the tirofiban and non-tirofiban groups.Differences in primary and safety outcomes between the matched groups were compared.Univariate and multivariate Logistic regression analyses were conducted with favorable outcomes at 90 days as the dependent variable to evaluate the effect of intravenous tirofiban administration after endovascular therapy on 90-day prognosis in patients without successful recanalization.Results A total of 356 patients without successful recanalization after endovascular treatment were included,comprising 195 males and 161 females,with a median age of 70(61,78)years(32-92 years).Among them,76 and 280patients were assigned to the tirofiban and non-tirofiban groups,respectively.At 90 days,114 patients had favorable outcomes,while 242 had unfavorable outcomes.(1)Before the 1∶1 propensity score matching,significant differences were observed between the tirofiban and non-tirofiban groups in terms of admission NIHSS score,ASPECTS,good collateral circulation,intravenous thrombolysis,and mTICI grade(all P＜0.05).After 1∶1 propensity score matching,66 matched pairs(132 patients)were obtained.No significant differences in the baseline and clinical characteristics were found between the two groups after matching(all P＞0.05).(2)After 1∶1 matching,a significant difference was observed in the rate of favorable outcomes at 90 days between the two groups(48.5％[32/66]vs.30.3％[20/66],P=0.033)after propensity score matching,while no significant differences were found in 90-day mortality,intracranial hemorrhage within 24h,or symptomatic intracranial hemorrhage within 24h(all P＞0.05).(3)Univariate analysis after 1∶1 matching indicated that age,atrial fibrillation,smoking history,admission NIHSS score,M2 segment middle cerebral artery occlusion,good collateral circulation,number of passes for thrombus retrieval,and tirofiban treatment were factors influencing favorable outcomes at 90 days.Multivariate Logistic regression analysis showed that younger age(aOR,0.942,95％CI 0.906-0.978,P=0.002),lower admission NIHSS score(aOR,0.855,95％CI 0.777-0.941,P=0.001),good collateral circulation(aOR,5.534,95％CI 2.141-14.301,P＜0.01),tirofiban treatment(aOR,2.774,95％CI 1.092-7.046,P=0.032),and M2 segment MCA occlusion(reference:internal carotid artery occlusion;aOR,4.874,95％CI 1.428-16.632,P=0.011)were independent predictors of favorable outcomes at 90 days.Conclusions Intravenous tirofiban administration after endovascular therapy may improve 90-day neurological outcomes in patients with acute anterior circulation large vessel occlusion stroke who did not achieve successful recanalization without increasing the risk of hemorrhage.Further large-scale randomized controlled trials are warranted to validate these findings.

Objective:To investigate the association between short-term blood pressure variability(BPV)and all-cause mortality in elderly patients with early sepsis-associated acute kidney injury(SA-AKI).Methods:A retrospective study was conducted on elderly patients with early SA-AKI from the Medical Information Mart for Intensive Care Ⅳ(MIMIC-Ⅳ)database between 2008 and 2022.All patients were divided into four groups(Q1-Q4)according to the quartiles of the 24-hour systolic blood pressure coefficient of variation(SBPCV): Q1(SBPCV<8.69%), Q2(8.69%≤SBPCV<10.84%), Q3(10.84%≤SBPCV<13.64%), Q4(SBPCV≥13.64%). Cox regression analysis was used to assess the association between SBPCV and one-year all-cause mortality.Restricted cubic spline regression was used for non-linear testing, and threshold effect analysis was further performed for non-linear relationships.Results:A total of 5 955 elderly patients were included in the study, and 2 746(46.11%)patients died within one year following the diagnosis of early SA-AKI.Patients in groups Q1 to Q4 had one-year all-cause mortality rates of 46.34%(690/1 489), 45.56%(678/1 488), 42.47%(632/1 488), and 50.07%(746/1 490), respectively.After adjusting for relevant confounding variables, Cox regression analysis showed that the Q1 group and Q4 group had a 26%( HR=1.26, 95% CI: 1.13-1.41, P<0.001)and 12%( HR=1.12, 95% CI: 1.01-1.24, P=0.047)higher risk of one-year mortality compared to the Q3 group, respectively.SBPCV and one-year all-cause mortality showed a U-shaped non-linear manner( P for nonlinear<0.001). Threshold effect analysis indicated that when SBPCV<13.07%, each standard deviation increase in SBPCV was associated with a 15% reduction in mortality risk( HR=0.85, 95% CI: 0.78-0.92, P<0.001); conversely, when SBPCV>13.07%, each standard deviation increase in SBPCV was associated with an 11% increase in mortality risk( HR=1.11, 95% CI: 1.03-1.21, P=0.011). Conclusions:Short-term BPV may be associated with long-term all-cause mortality in elderly patients with early SA-AKI in a U-shaped manner.However, these findings require further confirmation through high-quality prospective studies.

Objective To explore the application value of computed tomography(CT)based three-dimensional(3D)reconstruction technology in robot-assisted laparoscopic partial nephrectomy(RAPN)for patients with renal cell carcinoma(RCC).Methods A retrospective analysis was conducted on the clinical data of 185 RCC patients who underwent RAPN in our hospital during May 2021 and Aug.2023.Among them,98 underwent preoperative upper abdominal enhanced CT based 3D reconstruction(3D reconstruction group),and 87 did not perform it(non-3D reconstruction group).The differences in general information,tumor size,R.E.N.A.L.score,hospital stay,operation time,intraoperative blood loss,warm ischemia time,and postoperative complications were compared between the two groups.Results The 3D reconstruction group had clearer and more intuitive imaging.There were no significant differences in the general information,tumor size,R.E.N.A.L.score,intraoperative blood loss,and warm ischemia time between the two groups(P＞0.05).The 3D reconstruction group had shorter operation time[(125.8±34.9)min vs.(159.0±58.3)min],total hospital stay[(7.6±1.6)d vs.(7.7±3.0)d],and postoperative hospital stay[(3.7±0.5)d vs.(4.1±1.9)d],but longer preoperative hospital stay[(3.9±1.5)d vs.(3.5±2.2)d](all P＜0.001).Conclusion CT based 3D reconstruction technology can precisely display the anatomical parameters of tumor and its surrounding tissues,which can shorten the operation time,and postoperative and total hospital stay for RCC patients receiving RAPN.

Objective To identify key acoustic features associated with the progression of Alzheimer's disease(AD)through voice analysis combined with machine learning and feature selection techniques,thereby constructing classification models that serve as candidate tools for the early screening of AD.Methods Voice samples from AD,mild cognitive impairment(MCI)and healthy(HC)elderly individuals were obtained from the NCMMSC2021 AD voice dataset.The voice samples underwent data preprocessing,followed by feature extraction from the eGeMAPS feature set via the OpenSmile toolkit.Classification models were obtained utilizing random forest and support vector machine(SVM)algorithms.Significance testing and feature importance ranking were conducted using Python,and the further selection of the optimal features was performed through sequential forward selection(SFS).The classification performance before and after feature selection was compared and evaluated using accuracy and the area under the receiver operating characteristic curve(AUC).Results The significant acoustic features in the classification models primarily derived from spectral slope,formant,fundamental frequency,and loudness.The optimal classification performance was achieved with the SVM model following SFS feature selection,with recognition accuracies of 0.926(AUC=0.974)for AD/MCI group,0.875(AUC=0.956)for AD/HC group,and 0.879(AUC=0.904)for MCI/HC group.Conclusion SVM model performs better than random forest model,and the use of SFS for feature selection can effectively enhance model performance.Voice analysis has the potential to serve as a valuable supplementary tool for the rapid AD assessment and screening.