Development and validation of a simple model to predict functionally significant coronary artery disease in Chinese populations: A two-center retrospective study.

Objectives: This study sought to derive and validate a simple model combining traditional clinical risk factors with biomarkers and imaging indicators easily obtained from routine preoperative examinations to predict functionally significant coronary artery disease (CAD) in Chinese populations. Methods: We developed five models from a derivation cohort of 320 patients retrospective collected. In the derivation cohort, we assessed each model discrimination using the area under the receiver operating characteristic curve (AUC), reclassification using the integrated discrimination improvement (IDI) and net reclassification improvement (NRI), calibration using the Hosmer-Lemeshow test, and clinical benefit using decision curve analysis (DCA) to derive the optimal model. The optimal model was internally validated by bootstrapping, and external validation was performed in another cohort including 96 patients. Results: The optimal model including 5 predictors (age, sex, hyperlipidemia, hs-cTnI and LVEF) achieved an AUC of 0.807 with positive NRI and IDI in the derivation cohort. Moreover, the Hosmer-Lemeshow test showed a good fit, and the DCA demonstrated good clinical net benefit. The C-statistic calculated by bootstrapping internal validation was 0.798, and the calibration curve showed adequate calibration (Brier score = 0.179). In the external validation cohort, the optimal model performance was acceptable (AUC = 0.704; Brier score = 0.20). Finally, a nomogram based on this model was constructed to facilitate its use in clinical practice. Conclusions: A simple model combined clinical risk factors with hs-cTnI and LVEF improving the prediction of functionally significant CAD in Chinese populations. This attractive model may be a choice for clinicians to risk stratification for CAD.

Left atrial strain predicts risk and prognosis in patients with acute coronary syndrome: A retrospective study with external validation.

Objectives: To explore the association between left atrial (LA) strain and the GRACE score in patients with acute coronary syndrome (ACS) and to investigate the utility of LA function in predicting short-term adverse cardiovascular events post ACS. Methods: This retrospective study included ACS patients who underwent coronary angiography (CAG) in two independent cohorts from October 2020 to July 2022. The patients were classified into low-intermediate risk group and high-risk group based on the GRACE score. All participants underwent a transthoracic echocardiogram, with LA strain analysis before CAG. Correlation analysis was used to determine the relationship between LA strain and the GRACE score. The predictive value of LA strain was examined utilizing the area under the curve (AUC). Participants were followed for 10.5 ± 2.9 months for the primary endpoint of major adverse cardiovascular events (MACE). Results: A total of 229 patients were included in this study, including 196 in the primary group and 33 in the validation group. Spearman's correlation analysis showed there was a moderate negative correlation between the GRACE and left atrial reservoir strain (LASr) in both the primary (r = -0.63, P < 0.001) and validation (r = -0.73, P < 0.001) cohorts. Receiver operator characteristic (ROC) curve analysis showed that the AUC of LASr for prediction of the high-risk group was 0.86. Taking LASr 19.6% as the cut-off value, the sensitivity and specificity were 0.71 and 0.92, respectively. The cut-off value of 19.6% remains good at identifying high-risk group in the validation group (AUC = 0.87, sensitivity: 77.8%, specificity: 95.8%). Furthermore, 49 patients reached the endpoint in the primary cohort during the follow-up. On multivariable regression analysis, LASr (P = 0.03) was the independent echocardiographic predictor for the primary endpoint, rather than left atrial volume index (LAVI). Conclusions: LASr can identify high-risk patients with ACS as defined by the GRACE score and may be superior to Max LAVI in predicting incidents of MACE in the short-term following ACS.

Evaluation of longitudinal trajectory of functional tricuspid regurgitation on the risk of right ventricular dysfunction after mitral valve replacement.

OBJECTIVE: Functional tricuspid regurgitation (FTR) levels can vary over time and its longitudinal changing patterns may predict right ventricular dysfunction (RVD) risk. We aim to identify different trajectories of FTR in those who received mitral valve replacement (MVR) and investigate the association between longitudinal trajectory groups and RVD risk in a cohort study. METHODS AND RESULTS: A prospective cohort study, reported usual FTR levels at baseline in 2005-2015 and the participants of MVR have been followed up for 5-6 years, approximately every 1 year, and so far, the data have been collected across five subsequent phases. Five-year longitudinal trajectories of FTR were identified using group-based trajectory modeling (GBTM). We identified 3 distinct trajectories using a GBTM, labeled by initial value and changing pattern: stable group (258/378, 68.2%), increasing-slow group (67/378, 17.6%) and increasing-fast group (53/378, 14.2%). Treating the stable group as the reference, the age- and sex-adjusted odds ratio (OR) was 25.84 (95% confidence interval [CI]: 11.78-56.65) for the increasing-slow group and 139.94 (95% CI: 45.47-430.68) for the increasing-fast group by logistic regression model. After adjustment for every potential confounding factors, the OR is 14.21 (95% CI: 4.36-46.33) and 49.34 (95% CI: 8.88-273.87), respectively. CONCLUSIONS: The longitudinal trajectories of worsening FTR were mostly associated with increased risk of RVD outcomes, which is independent of other factors including FTR levels. These findings have implications for intervention and prevention of RVD among individuals who received MVR.

Using an Improved Residual Network to Identify PIK3CA Mutation Status in Breast Cancer on Ultrasound Image.

Background: The detection of phosphatidylinositol-3 kinase catalytic alpha (PIK3CA) gene mutations in breast cancer is a key step to design personalizing an optimal treatment strategy. Traditional genetic testing methods are invasive and time-consuming. It is urgent to find a non-invasive method to estimate the PIK3CA mutation status. Ultrasound (US), one of the most common methods for breast cancer screening, has the advantages of being non-invasive, fast imaging, and inexpensive. In this study, we propose to develop a deep convolutional neural network (DCNN) to identify PIK3CA mutations in breast cancer based on US images. Materials and Methods: We retrospectively collected 312 patients with pathologically confirmed breast cancer who underwent genetic testing. All US images (n=800) of breast cancer patients were collected and divided into the training set (n=600) and test set (n=200). A DCNN-Improved Residual Network (ImResNet) was designed to identify the PIK3CA mutations. We also compared the ImResNet model with the original ResNet50 model, classical machine learning models, and other deep learning models. Results: The proposed ImResNet model has the ability to identify PIK3CA mutations in breast cancer based on US images. Notably, our ImResNet model outperforms the original ResNet50, DenseNet201, Xception, MobileNetv2, and two machine learning models (SVM and KNN), with an average area under the curve (AUC) of 0.775. Moreover, the overall accuracy, average precision, recall rate, and F1-score of the ImResNet model achieved 74.50%, 74.17%, 73.35%, and 73.76%, respectively. All of these measures were significantly higher than other models. Conclusion: The ImResNet model gives an encouraging performance in predicting PIK3CA mutations based on breast US images, providing a new method for noninvasive gene prediction. In addition, this model could provide the basis for clinical adjustments and precision treatment.

Establishment and Validation of a Predictive Model for Long-Term Severe Functional Tricuspid Regurgitation after Mitral Valve Replacement.

BACKGROUND: The objective was to develop and validate an individualized nomogram to predict severe functional tricuspid regurgitation (S-FTR) after mitral valve replacement (MVR) via retrospective analysis of rheumatic heart disease (RHD) patients' pre-clinical characteristics. METHODS: Between 2001-2015, 442 MVR patients of RHD were examined. Transthoracic echocardiography detected S-FTR, and logistic regression model analyzed its independent predictors. R software established a nomogram prediction model, and Bootstrap determined its theoretical probability, which subsequently was compared with the actual patient probability to calculate the area under the curve (AUC) and calibration plots. Decision curve analysis (DCA) identified its clinical utility. RESULTS: Ninety-six patients developed S-FTR during the follow-up period. Both uni- and multivariate analyses found significant correlations between S-FTR occurrence with gender, age, atrial fibrillation (AF), pulmonary arterial hypertension (PH), left atrial diameter (LAD), and tricuspid regurgitation area (TRA). The individualized nomogram model had the AUC of 0.99 in internal verification. Calibration test indicated high agreement of predicted and actual S-FTR onset. DCA also showed that utilization of those six aforementioned factors was clinically useful. CONCLUSION: The nomogram for the patient characteristics of age, gender, AF, PH, LAD, and TRA found that they were highly predictive for future S-FTR onset within 5 years. This predictive ability therefore allows clinicians to optimize postoperative patient care and avoid unnecessary tricuspid valve surgeries.

Automatic myocardial infarction detection in contrast echocardiography based on polar residual network.

PURPOSE: Heart disease is one of the leading causes of death. Among patients with cardiovascular diseases, myocardial infarction (MI) is the main reason. Precise and timely identification of MI is significant for early treatment. Myocardial contrast echocardiography (MCE) is widely used for the detection of MI in clinic practice. However, existing clinical exam using MCE is subjective and highly operator dependent and time-consuming. Hence an automatic computer-aided MI detection in MCE is necessary to improve the diagnosis performance and decrease the workload of clinicians. METHODS: In this study, a novel deep learning model, polar residual network (PResNet) is proposed to identify MI regions in MCE images which design a polar layer considering the ring shape of the myocardium. MCE images are fed into the PResNet and a newly defined polar layer is used to describe the myocardium with a ring shape. The whole polar images are evenly divided into several subsections and a residual network is improved to classify the subsection into normal and abnormal categories. Finally, the detection results are mapped back to the original image to illustrate the infarction regions' locations for the further process. RESULTS: To evaluate the proposed PResNet, a dataset is constructed via performing MCE on five mice, which underwent the left anterior descending artery ligation and receive erythropoietin or saline injection, and the area variation fraction is manually annotated by an experienced expert as golden standards. The results demonstrate that the proposed PResNet model accomplishes high classification precisions with 99.6% and 98.7%, and 0.999 and 0.996 of AUC (area under the receiver operator curve) values on two different testing sets, respectively. Results suggest that the proposed model could enable accurate infarct detection and diagnosis of the MCE images. CONCLUSION: Those efficiency gains highlight the powerful ability to describe and interpret the MCE images using the polar layer and residual network. The proposed PResNet might aid the clinicians in fast and accurate assessing the infarcted myocardium on MCE.

[Primary pediatric central nervous system tumors statistic: study of 763 cases in a single institution].

OBJECTIVE: To investigate the epidemiology of primary pediatric central nervous system tumors. METHODS: The clinical data of 763 pediatric patients aged under 18 diagnosed as with primary central nervous system tumors in the past 10 years in Huashan Hospital were analyzed retrospectively. RESULTS: The mean age was 12.68 and the male to female ratio was 1.56:1. Among the tumors intracranial tumors accounted for 93.4% (713/763), whereas spinal tumors accounted for 6.6% (50/763). 476 of the 763 tumors (62.4%) were supratentorial, including pineal-quadrigeminal tumors, which was predominant to infratentorial tumors (30.5%, 233/763). 688 of the 763 patients had pathological diagnosis. The most common 5 categories of tumors were: astrocytic tumors (25.7%, 177/688), craniopharyngioma (12.8%, 88/688), medulloblastoma (9.3%, 64/688), germ cell tumors (8.9%, 61/688), and pituitary adenoma (7.5%, 52/688). CONCLUSION: As the main neurological medical center in the southern part of China, the statistics of Huashan Hospital can be representative. The epidemiology of pediatric central nervous system tumors has its own specialty. Therefore, it is crucial to deal with primary pediatric central nervous system tumors according to children's characteristics.