Diagnosis of congenital heart disease using Deep Learning in pediatric chest X-rays: A proof of concept
Main Article Content
Keywords
Heart Defects, Congenital, Artificial Intelligence, Radiography, Thoracic, Convolutional Neural Network, Precision Medicine
Abstract
Background: In Mexico, congenital heart diseases (CHD) are the most common birth defects. Despite their high mortality rate, many CHD are not detected early by general practitioners and pediatricians at the primary care level. Echocardiography, the diagnostic standard for CHD, is only available at tertiary care facilities with specialized equipment and personnel. In contrast, chest X-rays are an inexpensive and accessible test, and their analysis using artificial intelligence (AI) could allow for the presumptive diagnosis of CHD.
Objective: Create a database of pediatric chest X-rays and develop a proof of concept that evaluates the feasibility of applying AI algorithms for the presumptive detection of CHD from these images.
Material and methods: A retrospective cross-sectional study was conducted based on the analysis of pediatric chest X-ray images using a deep convolutional neural network implemented under a Residual Network (ResNet) architecture.
Results: Among the radiographs included in the study, 426 (65%) corresponded to patients with CC and 230 (35%) to patients without CC. A deep learning algorithm for binary classification (Healthy/Cardiopath) applied to this set achieved a diagnostic precision of 75%.
Conclusions: According to available records, this database represents the largest collection of chest X-rays from Mexican pediatric patients with CHD confirmed by clinical experts. This resource enabled the training of an AI-based model with sufficient diagnostic performance to support its potential utility as a presumptive screening tool in healthcare settings with limited access to pediatric cardiology subspecialists.
References
1. Liu Y, Chen S, Zühlke L, et al. Global birth prevalence of congenital heart defects 1970–2017: updated systematic review and meta-analysis of 260 studies. Int J Epidemiol. 2019;48(2):455–63. doi:10.1093/ije/dyz009.
2. Instituto Mexicano del Seguro Social. Lineamiento técnico médico para la realización del tamiz neonatal cardiaco [Internet]. Ciudad de México: IMSS; 2022 [consultado 3 de octubre de 2025]. Disponible en: https://www.gob.mx/cms/uploads/attachment/file/802995/Lineamiento_tamiz_neonatal_cardiaco_Aportaciones_Comite___2022_.pdf
3. Torres-Cosme JL, Rolón-Porras C, Aguinaga-Ríos M, et al. Mortality from congenital heart disease in Mexico: a problem on the rise. PLoS One. 2016;11(3):e0150422. doi:10.1371/journal.pone.0150422.
4. CENETEC. Diagnóstico y Tratamiento de Los Defectos de Los Tabiques Interauricular e Interventricular En Menores de 18 Años En Segundo y Tercer Nivel de Atención. Guía de Evidencias y Recomendaciones: Guía de Práctica Clínica, 2018 [citado 3 de octubre de 2025].
5. Baumgartner H, De Backer J, Babu-Narayan SV, et al. 2020 ESC guidelines for the management of adult congenital heart disease. Eur Heart J. 2021;42(6):563–645. doi:10.1093/eurheartj/ehaa554.
6. Murni IK, Wirawan MT, Patmasari L, et al. Delayed diagnosis in children with congenital heart disease: a mixed-method study. BMC Pediatr. 2021;21(1):191. doi:10.1186/s12887-021-02667-3.
7. Siregar MHR, Daulay ER, Hasan R. Role of chest radiograph in diagnosis of acyanotic congenital heart disease. J Soc Med. 2022;1(3):87–92.
8. Sarıoğlu T, Erek E, Yalçınbaş YK. Why national databases are necessary for pediatric and congenital heart surgery practice? Turk Gogus Kalp Damar Cerrahisi Derg. 2019;27(3):418–22. doi:10.5606/tgkdc.dergisi.2019.17945.
9. Markham JL, Hall M, Stephens JR, et al. A pediatric hospital medicine primer for performing research using administrative data. Hosp Pediatr. 2022;12(9):e319–25. doi:10.1542/hpeds.2022-006691.
10. Han PL, Jiang L, Cheng JL, et al. Artificial intelligence-assisted diagnosis of congenital heart disease and associated pulmonary arterial hypertension from chest radiographs: a multi-reader multi-case study. Eur J Radiol. 2024;171:111277. doi:10.1016/j.ejrad.2023.111277.
11. Chen R, Wu X, Chung T, et al. Using deep learning to detect pediatric congenital heart disease by chest radiography [preprint]. 2021 [citado 25 de septiembre de 2025]. doi:10.21203/rs.3.rs-491844/v1.
12. Zhixin L, Gang L, Zhixian J, et al. CHD-CXR: a de-identified publicly available dataset of chest x-ray for congenital heart disease. Front Cardiovasc Med. 2024;11:1351965. doi:10.3389/fcvm.2024.1351965.
13. Rajpurkar P, Irvin J, Zhu K, et al. CheXNet: radiologist-level pneumonia detection on chest X-rays with deep learning [preprint]. arXiv:1711.05225. 2017 [consultado 25 de septiembre de 2025]. Disponible en: https://arxiv.org/abs/1711.05225.
14. Giełczyk A, Marciniak A, Tarczewska M, Lutowski Z. Pre-processing methods in chest X-ray image classification. PLoS One. 2022;17(4):e0265949. doi:10.1371/journal.pone.0265949.
15. Tomás-Alvarado E, Ramírez-Terán OA. Descripción epidemiológica de niños con cardiopatías congénitas en Quintana Roo, México. Rev Mex Pediatr. 2024;91(2):64–7. doi:10.35366/119128.
16. Toba S, Mitani Y, Yodoya N, et al. Prediction of pulmonary to systemic flow ratio in patients with congenital heart disease using deep learning-based analysis of chest radiographs. JAMA Cardiol. 2020;5(4):449–57. doi:10.1001/jamacardio.2019.5620.
17. Luo L, Yu L, Chen H, et al. Deep mining external imperfect data for chest X-ray disease screening. IEEE Trans Med Imaging. 2020;39(11):3583–94. doi:10.1109/TMI.2020.3000949.
18. Rao PS. Diagnosis and management of ventricular septal defects. Rev Cardiovasc Med. 2024;25(11):411. doi:10.31083/j.rcm2511411.
19. Schiavone WA, Majdalany DS. Chest radiography pearls in select adult congenital heart disease. J Pers Med. 2024;14(4):397. doi:10.3390/jpm14040397.
20. Contreras-Hernández I, Bonifaz-Alfonzo LC. Redes transversales de investigación en salud: estrategia para la medicina traslacional. Rev Med Inst Mex Seguro Soc. 2022;60(Suppl 2):S110–8. doi: https://doi.org/10.5281/
21. Jiang F, Jiang Y, Zhi H, et al. Artificial intelligence in healthcare: past, present and future. Stroke Vasc Neurol. 2017;2(4):230–43. doi:10.1136/svn-2017-000101.
22. Yahya D, Stoyanova M, Hachmeriyan M, et al. The application of the facial analysis program Face2Gene in a single genetic counseling center: a retrospective study. Egypt Pediatr Assoc Gaz. 2025;73:1. doi:10.1186/s43054-025-00344-z.
23. Shroyer S, Mehta S, Thukral N, et al. Accuracy of cath lab activation decisions for STEMI-equivalent and mimic ECGs: physicians vs AI (Queen of Hearts by PMcardio). Am J Emerg Med. 2025;97:193–9. doi:10.1016/j.ajem.2025.07.061.
24. Esteva A, Robicquet A, Ramsundar B, et al. A guide to deep learning in healthcare. Nat Med. 2019;25(1):24–9. doi:10.1038/s41591-018-0316-z.
25. Garrido-Pino CA, López-Montero LM, López-Lozano L, et al. Edema macular en fotografías de fondo de ojo mediante un algoritmo computacional. Rev Med Inst Mex Seguro Soc. 2024;62(2):e5701. doi:10.5281/zenodo.10711610.
26. Sharma M, Savage C, Nair M, Larsson I, Svedberg P, Nygren JM. Artificial Intelligence Applications in Health Care Practice: Scoping Review. J Med Internet Res. 2022 Oct 5;24(10):e40238. doi: 10.2196/40238. PMID: 36197712; PMCID: PMC9582911.
27. Alowais SA, Alghamdi SS, Alsuhebany N, Alqahtani T, Alshaya AI, Almohareb SN, et al. Revolutionizing healthcare: the role of artificial intelligence in clinical practice. BMC Med Educ. 22 de septiembre de 2023;23(1):689.
28. Chávez-Martínez O, Ragacini LA. Educación médica e inteligencia artificial: perspectivas y desafíos éticos. Rev Med Inst Mex Seguro Soc. 2025;63(5):e6736. doi:10.5281/zenodo.16748310.
Article Sidebar
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
Authors retain their copyright and grant the Revista Médica del IMSS the right of first publication. Articles are distributed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which allows sharing as long as the author and the original source are properly credited.