Artificial Intelligence-enabled ECG Detection of Congenital Heart Disease in Children: a Novel Diagnostic Tool

Xinhua Hospital, Shanghai Jiao Tong University School of Medicine

Overview

Congenital heart disease (CHD) is the most common congenital disease in children. The early detection, diagnosis and treatment of CHD in children is of great significance to improve the prognosis and reduce the mortality of children, but the current screening methods have limitations. Electrocardiogram (ECG), as an economical and rapid means of heart disease detection, has a very important value in the auxiliary diagnosis of CHD.Big data and deep learning technologies in artificial intelligence (AI) have shown great potential in the medical field. The advent of the big data era provides rich data resources for the in-depth study of CHD ECG signals in children. The development of deep learning technology, especially the breakthrough in the field of image recognition, provides a strong technical support for the intelligent analysis of electrocardiogram. The particularity of children electrocardiogram requires the development of a special algorithm model. At present, the research on the application of deep learning models to identify children's electrocardiograms is limited, and the training and verification from large data sets are lacking. Based on the Chinese Congenital Heart Disease Collaborative Research Network, this project aims to integrate data and deep learning technology to develop a set of intelligent electrocardiogram assisted diagnosis system (CHD-ECG AI system) suitable for children with CHD, so as to improve the early detection rate of CHD and improve the efficiency of congenital heart disease screening.

Study Type

OBSERVATIONAL

Conditions

Artificial Intelligence Electrocardiogram Deep Learning Congenital Heart Disease in Children

Eligibility

Sex: ALLMin age: 3 MonthsMax age: 18 YearsHealthy volunteers:

Medical Language ↔ Plain English

Inclusion Criteria: * The age of first visit was from 3 months after birth to 18 years old; * In the atrial septal defect group, patients in the case group were required to complete ECG examination and confirmed by careful cardiac ultrasonography that there was a simple secondary atrial septal defect without other complex heart malformations (such as ectopic pulmonary vein drainage, trunk conus artery malformation, interrupted aortic arch, primary pulmonary hypertension, etc.). In the pulmonary hypertension group, the presence of CHD associated pulmonary hypertension was confirmed by careful cardiac ultrasonography examination. The control group was the patients with normal intracardiac structure examined by cardiac ultrasonography. The time interval between ECG examination and echocardiography examination of all patients was \< 1 month; * No major illness at the time of initial visit (non-life-threatening organic disease caused by congenital heart disease). Exclusion Criteria: * Age of first visit \< 3 months or \> 18 years old; * Complicated congenital heart disease (such as anomalous pulmonary venous drainage, trunk conus artery malformation, interrupted aortic arch, primary pulmonary hypertension, etc.); * The clinical information is incomplete, including the lack of ECG or echocardiography information, or the time interval between ECG and echocardiography is \> 1 month; * Life-threatening diseases associated with other organ systems;

Outcomes

Primary Outcomes

Large-scale ECG database for children

The ECG data of children from multiple centers were collected and collated, including common and rare CHD types and normal children's ECG, to construct a large-scale ECG database covering different ages and CHD diseases. In addition, the original ECG data (digital signals or ECG images) will be pre-processed to make it conform to the input standards of deep learning models, so as to improve the quality and efficiency of subsequent model training and reduce the heterogeneity of multi-center ECG data.

Time frame: 2024.01.01-2024.12.30

Artificial intelligence-assisted electrocardiogram model for CHD in Children

The deep neural network model will be established based on algorithms such as convolutional neural network, transformers and Autoencoders, and will be trained and verified in the multi-center children's ECG dataset (85%) established based on CCHDnet, so as to continuously optimize the model and improve the diagnostic performance of the model. Further, the deep learning model based on the single disease of CHD will be integrated, and the CHD-ECG AI system will be built, and the model will eventually automatically extract and recognize the general basic information such as the age and gender of the child through the ECG, and then predict and classify the potential CHD characteristics in the ECG based on this. The research group initially selected the representative subtypes of CHD - atrial septal defect and pulmonary hypertension as the initial direction of exploration.

Time frame: 2024.01.01-2025.12.30