Immune checkpoint inhibitors (ICIs) occupy a core position in tumor immunotherapy, and greatly improve the prognosis of tumor patients. However, with the deepening of clinical practice, it has been found that ICIs can not only activate the immune system, but also cause many adverse reactions, especially leading to a variety of cardiovascular toxicity and increased cardiovascular mortality. This paper reviews the clinical manifestations, pathogenesis, diagnosis and treatment, etc. of ICIsinduced cardiovascular toxicity, which provides evidences for preventing cardiovascular injury and improving cardiac function.

This article focuses on the application of artificial intelligence (AI) technology in the diagnosis and treatment of arrhythmias, and briefly introduces the latest relevant research findings from both domestic and international sources. It not only outlines the capability of AI in significantly improving the detection efficiency and accuracy of arrhythmia screening and diagnosis, but also emphasizes its value in the prediction of arrhythmia and the risk of its related complications, and guidance of its clinical treatment. Furthermore, this article summarizes the challenges faced by AI in arrhythmia management, including data quantity and quality issues, the “black box” effect in AI learning process, model generalization issues, ethical issues in technology, and the clinical acceptance of AI models. Finally, it looks forward to the prospect of AI technology deeply integrating with medicine, and comprehensive application in all aspects of arrhythmia diagnosis and treatment, so as to optimize medical resource allocation and medical service workflows, increase medical efficiency and quality, and improve patients’ prognoses.

Objective  To develop a transfer learning-based classification model for sleep apnea syndrome using electrocardiogram (ECG) data, increasing its classification accuracy and clinical applicability. Methods  Based on the Apnea-ECG and MIT-BIH polysomnographic databases, with respiratory signals as input, we applied a Butterworth low-pass filter for denoising, and constructed an original data set. To address the problem of insufficient respiratory signal data, a model training method based on a transfer learning approach was proposed: first, ECG signals with a large sample size were used for model pre-training, and then they were fine-tuning for respiratory signals, finally fulfilling binary classification or multi-class classification tasks. A cascade model combining residual network and bidirectional long short-term memory network was proposed, which performed better in capturing the timing features of signals and improving classification performance. Additionally, the performance of this model was made comparative analysis with those of various classic convolutional neural networks. Results  Through comparative experiments, it was found that employing transfer learning approach could accelerate model convergence and improve the model's overall performance. Validated on the test set, the proposed cascade model demonstrated a favorable performance in both binary classification and multi-class classification tasks, achieving an accuracy of 95.43% on the binary classification task and 91.26% on the multi-class classification task. Conclusion  This study offers novel insights into the design of disease classification models under small-sample conditions, and validates the effectiveness of transfer learning in sleep apnea syndrome classification, thereby demonstrating its potential clinical utility.

Objective  To explore the association between obstructive sleep apnea syndrome (OSAS) and heart rate variability (HRV) or circadian rhythm of blood pressure in hypertensives. Methods  A total of 102 inpatients and outpatients with hypertension were selected in our study, and they were divided into hypertension with OSAS group (55 cases) and primary hypertension group (47 cases). This study enrolled 50 concurrent healthy subjects (all with organic diseases excluded, and without comorbidities of hypertension or OSAS) as a control group. We compared and analyzed the HRV time-domain indexes of SDANN and rMSSD, and HRV frequency-domain indexes of LF and HF, as well as apnea hypopnea index (AHI), nocturnal mean oxygen saturation level, daytime and nocturnal blood pressure levels, and changes of circadian blood pressure rhythm. Results  The nocturnal minimum saturation of arterial oxygen (SaO2) and nocturnal mean SaO2 of patients in the hypertension with OSAS group were both lower than those of patients in the primary hypertension group, while AHI was higher (all P＜0.05). There were no statistically significant differences in the above sleep respiratory parameters between the primary hypertension group and the control group (all P＞0.05). There were no statistically significant differences in the daytime and nocturnal blood pressure levels, and the incidence of non-dipper pattern between the primary hypertension group and the control group (all P＞0.05). Comparisons of the aforementioned parameters between the hypertension with OSAS group and the control group all demonstrated statistically significant differences (all P＜0.01). The hypertension with OSAS group exhibited a significantly higher prevalence of non-dipper circadian blood pressure rhythm compared to the other two groups (all P＜0.05). Compared to the other two groups, the values of SDANN and LF of patients in the hypertension with OSAS group were higher, while the values of rMSSD and HF were lower (all P＜0.05). There were no statistically significant differences in the LF, HF, SDANN and rMSSD values between the primary hypertension group and the control group (all P＞0.05). Conclusion  Patients with OSAS demonstrate increased nocturnal blood pressure level, enhanced sympathetic nerve tension and decreased vagus nerve tension, as well as increased risk of cardiovascular and cerebrovascular diseases.

Objective  To explore the relationship between epicardial fat thickness (EFT), serum secreted frizzled-related protein 5 (SFRP5), wingless-type MMTV integration site family member 5a (Wnt5a), and in-stent restenosis (ISR) in patients with acute ST-segment elevation myocardial infarction (STEMI) after percutaneous coronary intervention (PCI). MethodsA total of 128 STEMI patients who had undergone emergency PCI were enrolled. After a 12-month follow-up, they were divided into a non-ISR group (80 cases) and an ISR group (48 cases) based on coronary angiography results. The general data between the two groups were compared, as well as the EFT before discharge and 1 month after surgery, and the serum levels of SFRP5 and Wnt5a measured before surgery and 1 month postoperatively. The differences in the levels of these indicators among patients with varying degrees of coronary artery lesions were also examined. Pearson correlation analysis was used to evaluate the association between EFT and serum levels of SFRP5 and Wnt5a 1 month after surgery. Multivariate Logistic regression analysis was applied to identify the risk factors of ISR in STEMI patients after PCI. ROC curve analysis was used to assess the predictive value of EFT, and serum SFRP5 and Wnt5a levels for ISR. ResultsThe EFT values of the ISR group before discharge and 1 month after surgery were both greater than those of the non-ISR group (all P<0.01). The serum SFRP5 level of the ISR group was lower than that of the non-ISR group 1 month after surgery, while the serum Wnt5a level was higher (all P<0.01). Compared to pre-operative levels, serum SFRP5 levels increased 1 month after surgery in the non-ISR group but decreased in the ISR group (both P<005). Serum Wnt5a levels showed a mild increase 1 month after surgery compared to pre-operative levels in the non-ISR group, but increased significantly in the ISR group (both P<0.05). The serum SFRP5 level 1 month after surgery was significantly lower in the three-vessel disease group than that in the single- or two-vessel disease groups (P<0.05). EFT 1 month after surgery showed a negative correlation with SFRP5 and a positive correlation with Wnt5a. Both univariate Logistic regression analysis and Logistic regression analysis adjusted for factors such as age and BMI demonstrated that EFT, and serum SFRP5 and Wnt5a levels 1 month after surgery were independent influencing factors for ISR. ROC curve analysis revealed that the combination of these three markers yielded an AUC value of 0.906, which was higher than that of detection with any single indicator. Conclusion  Increased EFT, decreased serum SFRP5, and elevated serum Wnt5a 1 month after surgery are closely related to ISR in STEMI patients after PCI. Early detection of these markers facilitates the prediction of ISR, and the combined detection offers superior predictive value.