No abstract

Objective To explore whether the post‐left atrium space (PLAS) ratio would be useful for prenatal diagnosis of total anomalous pulmonary venous connection (TAPVC) using echocardiography and artificial intelligence. Methods We retrospectively included 642 frames of four‐chamber views from 319 fetuses (32 with TAPVC and 287 without TAPVC) in end‐systolic and end‐diastolic periods with multiple apex directions. The average gestational age was 25.6 ± 2.7 weeks. No other cardiac or extracardiac malformations were observed. The dataset was divided into a training set (n = 540; 48 with TAPVC and 492 without TAPVC) and test set (n = 102; 20 with TAPVC and 82 without TAPVC). The PLAS ratio was defined as the ratio of the epicardium‐descending aortic distance to the center of the heart‐descending aortic distance. Supervised learning was used in DeepLabv3+, FastFCN, PSPNet, and DenseASPP segmentation models. The area under the curve (AUC) was used on the test set. Results Expert annotations showed that this ratio was not related to the period or apex direction. It was higher in the TAPVC group than in the control group detected by the expert and the four models. The AUC of expert annotations, DeepLabv3+, FastFCN, PSPNet, and DenseASPP were 0.977, 0.941, 0.925, 0.856, and 0.887, respectively. Conclusion Segmentation models achieve good diagnostic accuracy for TAPVC based on the PLAS ratio.

In Federated Learning (FL), data communication among clients is denied. However, it is difficult to learn from the decentralized client data, which is under-sampled, especially for segmentation tasks that need to extract enough contextual semantic information. Existing FL studies always average client models to one global model in segmentation tasks while neglecting the diverse knowledge extracted by the models. To maintain and utilize the diverse knowledge, we propose a novel training paradigm called Federated Learning with Z-average and Cross-teaching (FedZaCt) to deal with segmentation tasks. From the model parameters’ aspect, the Z-average method constructs individual client models, which maintain diverse knowledge from multiple client data. From the model distillation aspect, the Cross-teaching method transfers the other client models’ knowledge to supervise the local client model. In particular, FedZaCt does not have the global model during the training process. After training, all client models are aggregated into the global model by averaging all client model parameters. The proposed methods are applied to two medical image segmentation datasets including our private aortic dataset and a public HAM10000 dataset. Experimental results demonstrate that our methods can achieve higher Intersection over Union values and Dice scores.

Background and Objectives:To analyze echocardiographic parameters of fetal large ventricular septal defect (VSD) and tetralogy of Fallot (TOF) in the context of multicenter data and big data analysis because these two diseases are often misdiagnosed in fetuses, and to find the key parameters for the differential diagnosis of these two diseases.Methods: A total of 305 cases of large VSD and 192 cases of TOF diagnosed by fetal echocardiography from August 2010 to July 2016 from the database of Beijing Key Laboratory of Fetal Heart Defects were analyzed. Quantile regression of the 48 echocardiographic parameters of the 6272 normal fetuses from seven Chinese medical institutions was performed to determine the Q-score. The forward selection method and the naive Bayesian classification method were used to analyze the core differential diagnostic variables of fetal TOF and VSD. Results:The Q-score of the internal diameter of the aorta (AO Q-score), the ratio of the diameter of the pulmonary artery to the internal diameter of the aorta (PA/AO), and the Q-score of the ratio of the diameter of the pulmonary artery to the internal diameter of the aorta (PA/AO Q-score) were key parameters for the differential diagnosis of fetal large VSD and TOF. PA/AO was the primary parameter, with an area under the receiver operating characteristic curve of 0.951. Conclusions: These findings provide a new method for the prenatal diagnosis of large VSD and TOF and a theoretical basis for the intelligent diagnosis of large VSD and TOF. K E Y W O R D S big data analysis, fetal echocardiography, fetal large ventricular septal defect, tetralogy of fallot | 621 LV et aL. How to cite this article: Lv J, Yang T, Gu X, et al. Differential diagnosis of fetal large ventricular septal defect and tetralogy of Fallot based on big data analysis. Echocardiography.