A Real-Time End-to-End Framework with a Stacked Model Using Ultrasound Video for Cardiac Septal Defect Decision-Making
2024

Real-Time Framework for Cardiac Septal Defect Diagnosis Using Ultrasound Videos

Sample size: 222 publication Evidence: high

Author Information

Author(s): Nurmaini Siti, Nova Ria, Sapitri Ade Iriani, Rachmatullah Muhammad Naufal, Tutuko Bambang, Firdaus Firdaus, Darmawahyuni Annisa, Islami Anggun, Mandala Satria, Partan Radiyati Umi, Arum Akhiar Wista, Bastian Rio, Li Yanfeng

Primary Institution: Intelligent System Research Group, Universitas Sriwijaya, Palembang, Indonesia

Hypothesis

Can a real-time end-to-end framework using deep learning improve the diagnosis of cardiac septal defects from ultrasound videos?

Conclusion

The proposed model achieved high accuracy and robust performance in diagnosing cardiac septal defects from ultrasound videos.

Supporting Evidence

  • The model achieved 95.86% accuracy, 96.82% sensitivity, and 98.74% specificity during validation.
  • During real-time testing, the model exhibited 97.17% accuracy, 95.80% sensitivity, and 98.15% specificity.
  • Only 3 out of 35 US videos were diagnosed incorrectly during testing.

Takeaway

This study created a smart system that helps doctors quickly and accurately find heart problems in kids using ultrasound videos.

Methodology

The study used a stacked Yolo model for real-time analysis of ultrasound videos, classifying normal and abnormal conditions and detecting cardiac defects.

Limitations

The dataset did not include defect sizes, and the study was retrospective with a limited number of patients.

Participant Demographics

The dataset included recordings from 61 patients with ASDs, 79 patients with VSDs, 7 patients with AVSDs, and 75 patients with normal conditions.

Digital Object Identifier (DOI)

10.3390/jimaging10110280

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication