An ensemble of neural networks provides expert-level prenatal detection of complex congenital heart disease
| العنوان: | An ensemble of neural networks provides expert-level prenatal detection of complex congenital heart disease |
|---|---|
| المؤلفون: | Yili Zhao, Jami C. Levine, Rima Arnaout, Anita J. Moon-Grady, Lara Curran, Erin Chinn |
| المصدر: | Nature medicine, vol 27, iss 5 Nat Med |
| بيانات النشر: | Springer Science and Business Media LLC, 2021. |
| سنة النشر: | 2021 |
| مصطلحات موضوعية: | 0301 basic medicine, Heart disease, Echocardiography, Three-Dimensional, Cardiovascular, Medical and Health Sciences, Congenital, 0302 clinical medicine, Pregnancy, Prenatal Diagnosis, Prenatal, Mass Screening, Heart Defects, Ultrasonography, Pediatric, screening and diagnosis, Artificial neural network, Ultrasound, Area under the curve, Heart, General Medicine, Thorax, Detection, Heart Disease, Echocardiography, Pregnancy Trimester, Second, 030220 oncology & carcinogenesis, Cardiology, Biomedical Imaging, Female, Pregnancy Trimester, 4.2 Evaluation of markers and technologies, Adult, Heart Defects, Congenital, medicine.medical_specialty, Biometry, Neural Networks, Immunology, Bioengineering, Sensitivity and Specificity, Article, Ultrasonography, Prenatal, General Biochemistry, Genetics and Molecular Biology, Computer, Young Adult, 03 medical and health sciences, Deep Learning, Fetus, Text mining, Internal medicine, medicine, Humans, business.industry, Myocardium, Second, medicine.disease, Ensemble learning, Confidence interval, 4.1 Discovery and preclinical testing of markers and technologies, 030104 developmental biology, Three-Dimensional, Congenital Structural Anomalies, Neural Networks, Computer, business |
| الوصف: | Congenital heart disease (CHD) is the most common birth defect. Fetal screening ultrasound provides five views of the heart that together can detect 90% of complex CHD, but in practice, sensitivity is as low as 30%. Here, using 107,823 images from 1,326 retrospective echocardiograms and screening ultrasounds from 18- to 24-week fetuses, we trained an ensemble of neural networks to identify recommended cardiac views and distinguish between normal hearts and complex CHD. We also used segmentation models to calculate standard fetal cardiothoracic measurements. In an internal test set of 4,108 fetal surveys (0.9% CHD, >4.4 million images), the model achieved an area under the curve (AUC) of 0.99, 95% sensitivity (95% confidence interval (CI), 84–99%), 96% specificity (95% CI, 95–97%) and 100% negative predictive value in distinguishing normal from abnormal hearts. Model sensitivity was comparable to that of clinicians and remained robust on outside-hospital and lower-quality images. The model’s decisions were based on clinically relevant features. Cardiac measurements correlated with reported measures for normal and abnormal hearts. Applied to guideline-recommended imaging, ensemble learning models could significantly improve detection of fetal CHD, a critical and global diagnostic challenge. Deep learning can facilitate identification of congenital heart disease from fetal ultrasound screening, a diagnosis that in clinical practice is often missed. |
| وصف الملف: | application/pdf |
| تدمد: | 1546-170X 1078-8956 |
| URL الوصول: | https://explore.openaire.eu/search/publication?articleId=doi_dedup___::cd5329f9e092bff44fb09edec3bdca23 https://doi.org/10.1038/s41591-021-01342-5 |
| حقوق: | OPEN |
| رقم الأكسشن: | edsair.doi.dedup.....cd5329f9e092bff44fb09edec3bdca23 |
| قاعدة البيانات: | OpenAIRE |
Find this article in full text from Springer Verlag. | تدمد: | 1546170X 10788956 |
|---|