Related Experiment Video
Updated: Jun 28, 2025
Fetal Mouse Cardiovascular Imaging Using a High-frequency Ultrasound 30/45MHZ System
Published on: May 5, 2018
CHD-CXR: a de-identified publicly available dataset of chest x-ray for congenital heart disease
Li Zhixin1, Luo Gang1, Ji Zhixian1
1Heart Center, Women and Children's Hospital, Qingdao University, Qingdao, China.
Insights
Early detection of congenital heart disease in children is crucial. A new dataset of chest X-rays and cardiac ultrasound reports aids deep learning models in identifying subtle indicators of this common birth defect.
Area of Science:
- Medical imaging
- Artificial intelligence in healthcare
- Pediatric cardiology
Background:
- Congenital heart disease (CHD) is a leading cause of infant mortality, often undetected early due to subtle radiographic signs and limited expert availability.
- Traditional diagnostic methods using chest X-rays for CHD have limitations in detecting subtle cardiac manifestations.
- Advancements in deep learning and computer vision offer potential for improved early detection of CHD through analysis of medical imaging.
Purpose of the Study:
- To address the scarcity of annotated medical imaging datasets for artificial intelligence (AI) development in pediatric cardiology.
- To create and release a comprehensive dataset of pediatric chest X-rays and associated cardiac ultrasound reports for CHD research.
- To facilitate the advancement of machine learning models for the early detection of congenital heart disease.
Main Methods:
- Curated a dataset of 828 DICOM chest X-ray files from pediatric patients diagnosed with congenital heart disease.
- Included corresponding cardiac ultrasound reports to provide detailed clinical information.
- Developed and evaluated a preliminary deep learning model using the dataset to assess its utility.
Main Results:
- The released dataset contains high-quality, expert-annotated medical images and reports, emphasizing complex structural characteristics.
- A preliminary deep learning model trained on the dataset achieved an area under the receiver operating characteristic curve (ROC) of 0.85.
- Demonstrated the dataset's potential to improve the accuracy of CHD detection in children.
Conclusions:
- The newly released dataset is a valuable resource for advancing AI in the diagnosis of congenital heart disease.
- Deep learning models show promise in detecting subtle radiographic signs of CHD, improving early diagnosis rates.
- Further research and development using this dataset can enhance diagnostic capabilities, particularly in underserved regions.
Abstract:
Congenital heart disease is a prevalent birth defect, accounting for approximately one-third of major birth defects. The challenge lies in early detection, especially in underdeveloped medical regions where a shortage of specialized physicians often leads to oversight. While standardized chest x-rays can assist in diagnosis and treatment, their effectiveness is limited by subtle cardiac manifestations. However, the emergence of deep learning in computer vision has paved the way for detecting subtle changes in chest x-rays, such as lung vessel density, enabling the detection of congenital heart disease in children. This highlights the need for further investigation. The lack of expert-annotated, high-quality medical image datasets hinders the progress of medical image artificial intelligence. In response, we have released a dataset containing 828 DICOM chest x-ray files from children with diagnosed congenital heart disease, alongside corresponding cardiac ultrasound reports. This dataset emphasizes complex structural characteristics, facilitating the transition from machine learning to machine teaching in deep learning. To ascertain the dataset's applicability, we trained a preliminary model and achieved an area under the receiver operating characteristic curve (ROC 0.85). We provide detailed introductions and publicly available datasets at: https://www.kaggle.com/competitions/congenital-heart-disease.
More Related Videos
11:09High-Resolution Cardiac Positron Emission Tomography/Computed Tomography for Small Animals
Published on: December 16, 2022
12:15Tissue Preparation Techniques for Contrast-Enhanced Micro Computed Tomography Imaging of Large Mammalian Cardiac Models with Chronic Disease
Published on: February 8, 2022
