Search

In this paper, we present the Multi-Center Privacy-Preserving Network (MP-Net), a novel framework designed for secure medical image segmentation in multi-center collaborations. Our methodology offers a new approach to multi-center collaborative learning, capable of reducing the volume of data transmission and enhancing data privacy protection. Unlike federated learning, which requires the transmission of model data between the central server and local servers in each round, our method only necessitates a single transfer of encrypted data. The proposed MP-Net comprises a three-layer model, consisting of encryption, segmentation, and decryption networks. We encrypt the image data into ciphertext using an encryption network and introduce an improved U-Net for image ciphertext segmentation. Finally, the segmentation mask is obtained through a decryption network. This architecture enables ciphertext-based image segmentation through computable image encryption. We evaluate the effectiveness of our approach on three datasets, including two cardiac MRI datasets and a CTPA dataset. Our results demonstrate that the MP-Net can securely utilize data from multiple centers to establish a more robust and information-rich segmentation model.

Background: The risk of coronary artery disease (CAD) in different valve dysfunction has been unclear., Methods: We reviewed patients, who underwent valve heart surgery and coronary angiography from 2008 to 2021, at our center., Results: A total of 7,932 patients were included in the present study, and 1,332 (16.8%) had CAD. The mean age of the study cohort was 60.5±7.9 years, and 4,206 (53.0%) were male. CAD was 21.4% in aortic disease, 16.2% in mitral valve disease, 11.8% in isolated tricuspid valve disease, and 13.0% in combined aortic and mitral valve disease. Patients with aortic stenosis were older than those with regurgitation (63.6±7.4 years vs. 59.5±8.2 years, P < 0.001), and the CAD risks also were higher (28.0% vs. 19.2%, P < 0.001). The age difference was minimal (60.6±8.2 years vs. 59.5±6.7 years, P = 0.002) between patients with mitral valve regurgitation and stenosis, but the risks of CAD were twice high in regurgitation (20.2% vs. 10.5%, P < 0.001). When the type of valve impairment was not considered, non-rheumatic etiology, advanced age, male sex, hypertension, and diabetes were independent predictors of CAD., Conclusion: In patients undergoing valve surgery, the prevalence of CAD was influenced by conventional risk factors. Importantly, CAD also was associated with the type and etiology of valve diseases.

Brain tumor segmentation remains a challenge in medical image segmentation tasks. With the application of transformer in various computer vision tasks, transformer blocks show the capability of learning long-distance dependency in global space, which is complementary to CNNs. In this paper, we proposed a novel transformer-based generative adversarial network to automatically segment brain tumors with multi-modalities MRI. Our architecture consists of a generator and a discriminator, which is trained in min-max game progress. The generator is based on a typical "U-shaped" encoder-decoder architecture, whose bottom layer is composed of transformer blocks with Resnet. Besides, the generator is trained with deep supervision technology. The discriminator we designed is a CNN-based network with multi-scale L 1 loss, which is proved to be effective for medical semantic image segmentation. To validate the effectiveness of our method, we conducted exclusive experiments on BRATS2015 dataset, achieving comparable or better performance than previous state-of-the-art methods. On additional datasets, including BRATS2018 and BRATS2020, experimental results prove that our technique is capable of generalizing successfully., Competing Interests: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest., (Copyright © 2022 Huang, Zhu, Chen, Wang, Chai and Zhang.)

Background: The factors that associated with the response of moderate ischemic mitral regurgitation (IMR) to isolated coronary artery bypass grafting (CABG) remain unclear. This study aims to evaluate whether left ventricular (LV) myocardial scar assessed by cardiovascular magnetic resonance (CMR) is associated with the outcome of moderate IMR after isolated CABG., Methods: Forty-six patients with coronary artery disease (CAD) and moderate IMR who underwent isolated CABG between January 2014 and February 2019 in Anzhen Hospital Affiliated to Capital Medical University were enrolled in this case-control study. All patients underwent CMR and echocardiography before surgery. Patients were classified into two groups according to the severity of IMR 1 year after CABG: an improved group (no or mild IMR) and an unimproved group (moderate or severe IMR). Univariate and multivariate logistic regression analyses were used to assess the association between individual variables and unimproved IMR at 1-year post-CABG., Results: Compared to patients in the improved group, the patients in the unimproved group had a significantly greater amount of LV myocardial scar (18.0%±9.5% vs. 30.8%±11.2%, P<0.001). In the multiple logistic regression model, after adjustment for age, sex, and body mass index, only LV myocardial scar (OR: 0.89, 95% CI: 0.83-0.96, P=0.001) was independently associated with unimproved IMR after isolated CABG. Furthermore, there was no difference in the 3-year overall survival rates between the two groups (92.3% vs. 90.3%, P=0.46). In addition, patients in the unimproved group had a higher New York Heart Association (NYHA) classification (P=0.01) and more major adverse cardiac events such as MI, angina pectoris, and readmission for heart failure (P=0.03)., Conclusions: A greater amount of preoperative LV myocardial scar was associated with unimproved moderate IMR after isolated CABG. Measuring preoperative LV myocardial scar is helpful to predict post-operative outcome and determine optimal surgery in patients with moderate IMR., Competing Interests: Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://dx.doi.org/10.21037/atm-21-3622). The authors have no conflicts of interest to declare., (2021 Annals of Translational Medicine. All rights reserved.)

The triglyceride-glucose index is a valuable marker of insulin resistance. However, the predictive value of this index for postoperative atrial fibrillation in patients undergoing septal myectomy remains unclear. A total of 409 patients with hypertrophic obstructive cardiomyopathy who underwent septal myectomy were recruited in this study. The triglyceride-glucose index was calculated for all patients preoperatively. All patients underwent clinical data collection, blood sampling, and standard echocardiographic examinations during hospitalization. The prevalence of postoperative atrial fibrillation was approximately 15% in the present study. Multivariate logistic regression revealed that age (odds ratio [OR]: 1.053, 95% CI: 1.016-1.090, P = 0.004), hypertension (OR: 2.399, 95% CI: 1.228-4.686, P = 0.010), left atrial diameter (OR: 1.101, 95% CI: 1.050-1.155, P < 0.001), and triglyceride-glucose index (OR: 4.218, 95% CI: 2.381-7.473, P < 0.001) were independent risk factors for postoperative atrial fibrillation in patients undergoing septal myectomy. In receiver operating characteristic curve analysis, the triglyceride-glucose index could provide a moderate predictive value for postoperative atrial fibrillation after septal myectomy 0.723 (95% CI: 0.650-0.796, P < 0.001). Moreover, adding the triglyceride-glucose index to conventional risk factor model could numerically but not significantly increase our ability to predict postoperative atrial fibrillation (area under the receiver: 0.742 (0.671-0.814) vs. 0.793 (0.726-0.860), p = 0.065) after septal myectomy. In our retrospective cohort study, the triglyceride-glucose index was identified as an independent predictor of postoperative atrial fibrillation in patients undergoing septal myectomy., Competing Interests: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest., (Copyright © 2021 Wei, Zhu, Ren, Dai, Li and Lai.)

Background: Previous studies have demonstrated that secreted frizzled-related protein 4 (SFRP4) is associated with impaired glucose and triglyceride metabolism in patients with stable coronary artery disease. In the present study, we investigated human epicardial adipose tissue (EAT)-derived and circulating SFRP4 levels in patients with coronary artery disease (CAD)., Methods: Plasma samples and adipose biopsies from EAT and subcutaneous adipose tissue (SAT) were collected from patients with CAD (n = 40) and without CAD (non-CAD, n = 30) during elective cardiac surgery. The presence of CAD was identified by coronary angiography. SFRP4 mRNA and protein expression levels in adipose tissue were detected by quantitative real-time PCR and immunohistochemistry, respectively. Plasma SFRP4 concentrations were measured by an enzyme-linked immunosorbent assay (ELISA). Correlation analysis and multivariate linear regression analysis were used to determine the association of SFRP4 expression with atherosclerosis as well as clinical risk factors., Results: SFRP4 mRNA and protein expression levels were significantly lower in EAT than in paired SAT in patients with and without CAD (all P < 0.05). Compared to non-CAD patients, CAD patients had higher SFRP4 expression levels in EAT (both mRNA and protein levels) and in plasma. Multivariate linear regression analysis showed that CAD was an independent predictor of SFRP4 expression levels in EAT (beta = 0.442, 95% CI 0.030-0.814; P = 0.036) and in plasma (beta = 0.300, 95% CI 0.056-0.545; P = 0.017). SAT-derived SFRP4 mRNA levels were independently associated with fasting insulin levels (beta = 0.382, 95% CI 0.008-0.756; P = 0.045). In addition, plasma SFRP4 levels were positively correlated with BMI (r = 0.259, P = 0.030), fasting insulin levels (r = 0.306, P = 0.010) and homeostasis model assessment of insulin resistance (HOMA-IR) values (r = 0.331, P = 0.005)., Conclusions: EAT-derived and circulating SFRP4 expression levels were increased in patients with CAD. EAT SFRP4 mRNA levels and plasma SFRP4 concentrations were independently associated with the presence of CAD.

Background: Omentin-1, a novel adipocytokine mainly expressed in visceral adipose tissue, has been found to inhibit the inflammatory response and improve insulin resistance as well as other obesity-related disorders. This study investigated the association between omentin-1 expression in human epicardial adipose tissue (EAT) and coronary atherosclerosis., Methods: Serum samples, and paired biopsies from EAT and subcutaneous adipose tissue (SAT), were obtained from patients with and without coronary artery disease (CAD, n = 28 and NCAD, n = 12, respectively) during elective cardiac surgery. Coronary angiography was performed to identify CAD presence. Serum omentin-1 and adiponectin levels were measured by ELISA. mRNA expression of omentin-1 and adiponectin was detected in adipose tissue by quantitative real-time PCR, and omentin-1 protein expression was evaluated by immunohistochemistry. Correlation and multivariate linear regression analyses were performed to determine the association between omentin-1 expression and clinical risk factors., Results: mRNA and protein expression of omentin-1 were higher in EAT than paired SAT in patients with CAD and NCAD. Compared with NCAD patients, CAD patients had lower omentin-1 and adiponectin mRNA levels in EAT and serum levels as well as lower omentin-1 protein levels. Among patients with CAD, omentin-1 expression was lower in EAT surrounding coronary segments with stenosis than those without stenosis, in terms of mRNA and protein, whereas adiponectin mRNA level in EAT did not seem to differ between stenotic and non-stenotic coronary segments in CAD patients. In multivariate linear regression analysis, CAD was an independent predictor of EAT omentin-1 mRNA expression (beta = -0.57, 95 % CI -0.89 to -0.24; P = 0.001) and serum omentin-1 levels (beta = -0.35, 95 % CI -0.67 to -0.03; P = 0.036)., Conclusions: Circulating and EAT-derived omentin-1 levels were reduced in patients with CAD. Omentin-1 expression in patients with CAD was lower in EAT adjacent to coronary stenotic segments than non-stenotic segments.

Objective: To evaluate the impact of preoperative dual antiplatelet therapy of aspirin and clopidogrel on perioperative blood loss and blood transfusion requirements., Methods: A total of 60 patients underwent off-pump coronary artery bypass grafting (OPCABG) performed by the same surgeons. And they were assigned to receive dual antiplatelet therapy of aspirin and clopidogrel (group AC, n = 30) and discontinue antiplatelet therapy more than 5 days before surgery (group control, n = 30)., Results: No significant inter-group differences existed in basic clinical characteristics (all P > 0.05). Itraoperative blood loss volume was similar for two groups (637 ± 257 vs 635 ± 196 ml, P = 0.978). No significant inter-group difference existed in 24 h chest drainage volume (522 ± 160 vs 524 ± 204 ml, P = 0.961) or total volume of chest drainage (1044 ± 337 vs 1118 ± 198 ml, P = 0.306)., Conclusion: Preoperative dual antiplatelet therapy of aspirin and clopidogrel does not increase the total volume of postoperative chest drainage.