Objective To systematically evaluate the incidence rate and predictors of early(≤48 h) malignant ventricular arrhythmia (MVA) in patients with acute myocardial infarction (AMI), and to establish and validate a clinical prediction model to assist in identifying high-risk patients. Methods The clinical data of 278 patients with AMI in the hospital from February 2023 to February 2025 were retrospectively analyzed. Based on MVA occurrence within 48 hours post-AMI, patients were divided into the MVA group (n=225 cases) and non-MVA group (n=53 cases).. The clinical data in the two groups were collected, and the predictive variables were determined by univariate Logistic analysis and multivariate Logistic regression analysis to establish a prediction model for MVA. Results The proportion of patients with Killip grade III or IV in MVA group was higher than that in non-MVA group (P<0.05), and the levels of white blood cell count (WBC), creatine kinase isoenzyme (CKMB) and troponin I (TnI) were also higher than those in non-MVA group (P<0.05) while the standard deviation of normal to normal RR intervals (SDNN) and left ventricular ejection fraction (LVEF) were lower than those in non-MVA group (P<0.05). Multivariate Logistic regression analysis showed that Killip grade≥III, high levels of WBC, CKMB and TnI and low SDNN and LVEF were independent risk factors of early MVA in patients with AMI (P<0.05). Based on the above six factors, a risk nomogram prediction model was constructed, and the model verification results showed that the area under the ROC curve (AUC) was 0.884 (95%CI: 0.835-0.932), with good model discrimination. The calibration curve was close to the ideal curve (Hosmer-Lemeshoe P=0.768), and the model had good predictive efficiency. The decision curve showed that the model had a higher predicted net benefit value (threshold=0.1-0.97). Conclusion Cardiac function Killip grade≥III, high WBC, CKMB and TnI and low SDNN and LVEF are independent risk factors of early MVA in AMI patients. The clinical prediction model based on the above variables has certain predictive value on the risk of MVA in AMI patients.

Objective:To develop a deep learning-based system for real-time recognition and classification of portal hypertensive gastropathy (PHG) and evaluate its ability to assist junior endoscopists.Methods:A total of 2 848 gastroscopy images from 832 patients with liver cirrhosis were selected from Digestive Endoscopy Center databases of Renmin Hospital of Wuhan University, Wuhan Hospital of Traditional Chinese and Western Medicine, and the Second Hospital of Jingzhou from January 2015 to October 2023. This system referred to 3 endoscopic features of Baveno Ⅱ scoring system. Three models were developed respectively for gastric antral vascular ectasia (GAVE), mosaic-like pattern (MLP), and red marks (RM). The specific classification references were as follows: (1) GAVE model: 0 no, 1 yes; (2) MLP model: 0 no, 1 mild, 2 severe; (3) RM model: 0 no, 1 isolated, 2 fused. The classification results for endoscopic characteristics of PHG of 3 endoscopy experts were taken as the gold standard. The yolov8-m model was used for training. The training dataset, validation dataset, and test dataset were allocated at a ratio of 8∶1∶1. The test dataset was used to evaluate the performance of models and their auxiliary effects on endoscopists. The accuracy, recall, precision, specificity and Kappa coefficient were calculated. Results:The accuracy, recall, specificity of GAVE model were 96.0% (48/50), 87.5% (7/8) and 97.6% (41/42). There was no significant difference between its accuracy and the gold standard ( χ2=316.226, P=1.000). The precision of GAVE1 and GAVE0 were 87.5% (7/8) and 97.6% (41/42) respectively. The accuracy of MLP model was 84.1% (132/157), and there was no significant difference compared with the gold standard ( χ2=3.286, P=0.193). The precision and recall of MLP2 were 88.2% (15/17) and 75.0% (15/20). The precision and recall of MLP1 were 77.9% (60/77) and 88.2% (60/68). The precision and recall of MLP0 were 90.5% (57/63) and 82.6% (57/69). The accuracy of RM model was 87.9% (123/140), and there was no significant difference compared with the gold standard ( χ2=2.891, P=0.409). The precision and recall of RM2 were 94.7% (18/19) and 78.3% (18/23). The precision and recall of RM1 were 72.2% (26/36) and 81.3% (26/32). The precision and recall of RM0 were 92.9% (79/85) and 92.9% (79/85). The mean accuracy of the three junior endoscopists, with and without the assistance of the GAVE model, MLP model, and RM model, respectively increased from 95.3% to 99.3%, from 83.9% to 91.9%, and from 81.9% to 83.1%. The overall consistency analysis of the 3 junior endoscopists with the gold standard indicated that the consistency of the GAVE model before and after assistance was extremely strong (both an overall Kappa of 1.000); the consistency before assistance of the MLP model was moderate (with an overall Kappa of 0.601), which increased to extremely strong after assistance (with an overall Kappa of 0.964); and the consistency of the RM model before and after assistance was also relatively strong (with an overall Kappa of 0.792 before and 0.798 after). Conclusion:The deep learning system accurately identifies and classifies PHG features and significantly enhances diagnostic performance of junior endoscopists.

Objective:To develop a deep learning-based system for real-time recognition and classification of portal hypertensive gastropathy (PHG) and evaluate its ability to assist junior endoscopists.Methods:A total of 2 848 gastroscopy images from 832 patients with liver cirrhosis were selected from Digestive Endoscopy Center databases of Renmin Hospital of Wuhan University, Wuhan Hospital of Traditional Chinese and Western Medicine, and the Second Hospital of Jingzhou from January 2015 to October 2023. This system referred to 3 endoscopic features of Baveno Ⅱ scoring system. Three models were developed respectively for gastric antral vascular ectasia (GAVE), mosaic-like pattern (MLP), and red marks (RM). The specific classification references were as follows: (1) GAVE model: 0 no, 1 yes; (2) MLP model: 0 no, 1 mild, 2 severe; (3) RM model: 0 no, 1 isolated, 2 fused. The classification results for endoscopic characteristics of PHG of 3 endoscopy experts were taken as the gold standard. The yolov8-m model was used for training. The training dataset, validation dataset, and test dataset were allocated at a ratio of 8∶1∶1. The test dataset was used to evaluate the performance of models and their auxiliary effects on endoscopists. The accuracy, recall, precision, specificity and Kappa coefficient were calculated. Results:The accuracy, recall, specificity of GAVE model were 96.0% (48/50), 87.5% (7/8) and 97.6% (41/42). There was no significant difference between its accuracy and the gold standard ( χ2=316.226, P=1.000). The precision of GAVE1 and GAVE0 were 87.5% (7/8) and 97.6% (41/42) respectively. The accuracy of MLP model was 84.1% (132/157), and there was no significant difference compared with the gold standard ( χ2=3.286, P=0.193). The precision and recall of MLP2 were 88.2% (15/17) and 75.0% (15/20). The precision and recall of MLP1 were 77.9% (60/77) and 88.2% (60/68). The precision and recall of MLP0 were 90.5% (57/63) and 82.6% (57/69). The accuracy of RM model was 87.9% (123/140), and there was no significant difference compared with the gold standard ( χ2=2.891, P=0.409). The precision and recall of RM2 were 94.7% (18/19) and 78.3% (18/23). The precision and recall of RM1 were 72.2% (26/36) and 81.3% (26/32). The precision and recall of RM0 were 92.9% (79/85) and 92.9% (79/85). The mean accuracy of the three junior endoscopists, with and without the assistance of the GAVE model, MLP model, and RM model, respectively increased from 95.3% to 99.3%, from 83.9% to 91.9%, and from 81.9% to 83.1%. The overall consistency analysis of the 3 junior endoscopists with the gold standard indicated that the consistency of the GAVE model before and after assistance was extremely strong (both an overall Kappa of 1.000); the consistency before assistance of the MLP model was moderate (with an overall Kappa of 0.601), which increased to extremely strong after assistance (with an overall Kappa of 0.964); and the consistency of the RM model before and after assistance was also relatively strong (with an overall Kappa of 0.792 before and 0.798 after). Conclusion:The deep learning system accurately identifies and classifies PHG features and significantly enhances diagnostic performance of junior endoscopists.

Objective:To serially characterize the myocardial perfusion, myocardial hibernation and left ventricular (LV) function as well as LV remodeling in progressive coronary artery stenosis in Chinese mini-pigs.Methods:In 8 Chinese mini-pigs (5 males, 3 females; age: 10 months), chronic progression of coronary stenosis and finally occlusion was established using Ameroid constrictor implantation at the 1 cm below the bifurcation of the first diagonal branch of the left anterior descending (LAD) artery. Serial gated 99Tc m-methoxyisobutylisonitrile(MIBI) SPECT/CT, gated 18F-FDG PET/CT imaging and coronary angiography (CAG) were performed before surgery and at the 1st, 4th and 8th week after surgery. Longitudinally, total perfusion defect (TPD), LV ejection fraction (LVEF), LV end-diastolic volume (LVEDV), LV end-systolic volume (LVESV), summed motion score (SMS), summed thickening score (STS) and hibernating myocardium (HM) were analyzed. Repeated measures analysis of variance, Kruskal-Wallis rank sum test and Bonferroni correction method were used to analyze data. Results:One mini-pig died of infection after the Ameroid constrictor implantation. In the remaining 7 mini-pigs, TPD was progressively increased with time prolonged (0, 12.0%(0, 33.0%), (41.1±23.7)% and (49.3±24.5)%; H=17.03, P=0.001); Compared with HM before the surgery (100%), HM was gradually reduced from the 1st (21.0%(6.0%, 100%)) to the 4th (18.0%(3.0%, 33.0%)) week after surgery, and then increased to the 8th week after surgery ((23.0±15.4)%; H=13.09, P=0.004), but there was no significant difference between the 1st and 4th week, or between the 4th and 8th week after surgery (both P>0.05 (Bonferroni correction method)). Accordingly, LVEF gradually decreased ((73.7±8.4)%, (63.7±19.1)%, (53.7±14.6)% and (49.9±15.4)%; F=6.22, P=0.004). LVEDV (9.0(6.0, 21.0), (31.4±16.3), (32.9±17.4) and (36.4±17.5) ml; H=8.58, P=0.035)and LVESV ((3.8±3.2), (15.9±15.3), 12.0(10.0, 17.0)and (19.3±10.9) ml; H=10.51, P=0.015) gradually increased. SMS and STS continuously increased as well ( H values: 16.49, 13.33, P values: 0.001, 0.004). Conclusions:With the progression of coronary artery stenosis to occlusion, myocardial perfusion is gradually decreased, while the global and regional LV function, LV remodeling are gradually aggravated, and HM is gradually reduced. After the chronic coronary artery occlusion, myocardial perfusion has a trend to be improved and HM is gradually recovered.

Objective:To evaluate the effect of remote ischemic conditioning (RIC) on left ventricular (LV) myocardial perfusion, myocardial viability, LV remodeling, regional and global LV function serially following acute myocardial infarction (AMI) in Chinese mini-pigs.Methods:AMI was established in 12 Chinese mini-pigs (8 males, 4 females; age: 6-8 months) and they were randomly divided into RIC group ( n=6) and non-RIC group ( n=6). RIC was performed in pigs by blood pressure inflation on the lower limbs for 5 min period and 4 cycles immediately after surgery. A series of myocardial perfusion imaging and gated 18F-fluorodeoxyglucose (FDG) myocardial metabolism PET/CT imaging were performed longitudinally at the 1st, 14th, 28th and 56th days after AMI, and parameters including total perfusion defect (TPD), hibernating myocardium (HM), Scar, left ventricular ejection fraction (LVEF), left ventricular end-diastolic volume (LVEDV), left ventricular end-systolic volume (LVESV), summed motion score (SMS), summed thickening score (STS) and changes of these parameters were obtained. Independent-samples t test and Mann-Whitney U test were used to analyze data. Results:Nine Chinese mini-pigs survived after surgery and were performed imaging. Compared to non-RIC group ( n=4), HM at the 28th ((6.0±2.4)% vs (17.0±4.6)%; t=-4.158), TPD 14th-1st ((-11.8±5.4)% vs 9.0%(4.5%, 15.0%); z=2.449), TPD 28th-1st ((-15.3±3.9)% vs (12.0±3.0)%; t=-10.071), TPD 56th-1st ((-18.0±6.5)% vs 9.0%(4.5%, 12.0%); z=2.449), HM 28th-1st ((-10.5±6.9)% vs (8.3±2.1)%; t=-4.507), HM 56th-1st (-15.0%(-17.5%, -8.5%) vs 2.0%(0%, 7.0%); z=2.449) and LVEDV 14th-1st (-0.5(-2.5, 0) ml vs (13.0±4.4) ml; z=2.470) were reduced in RIC group ( n=5; all P<0.05). Conclusion:RIC can improve myocardial perfusion, delay LV remodeling in the acute stage and salvage hibernating myocardium in the subacute stage and chronic stage.

CAR-T cell therapy that targets surface antigens to kill tumor cells specifically has recently become another cornerstone in tumor immunotherapy. In this study, a lentiviral expression plasmid of CAR targeting human epidermal growth factor receptor 2 (HER2) was constructed by genetic engineering. The recombinant plasmid was co-transfected with other packaging plasmids into HEK293T cells by calcium phosphate precipitation to generate lenti-car, which are CAR lentiviral particles. HER2-specific CAR-T cells were obtained by transducing human peripheral blood mononuclear cells with lenti-car. Their specific inhibitory effects on HER2-positive and HER2-negative tumor cells were analyzed in vitro. The constructed CAR-T cells were specifically activated by HER2-expressing tumor cells as indicated by secretion of IFN-γ and IL-2. The inhibitory rate on HER2-positive SK-OV-3 cell line was (58.47±1.72)%, significantly higher than that on the mock-treated control group (P<0.05). The inhibitory rate on HER2-negative K562 cell lines was (11.74±2.37)%, which was not significantly different from that on the control group (P>0.05). Furthermore, when we transfected a HER2-expressing vector into K562, the inhibitory rate increased to (30.41±7.59)%, which was higher than that on HER2-negative K562 (P<0.05). Thus, the constructed second-generation HER2-specific CAR-T cells specifically suppressed growth of tumor cells overexpressing HER2 protein, suggesting that HER2-specific CAR-T cells might prove useful for immunotherapy of HER2-positive cancer.