Related Experiment Video
Updated: May 13, 2025
07:34
Fetal Mouse Cardiovascular Imaging Using a High-frequency Ultrasound 30/45MHZ System
Published on: May 5, 2018
11.5K
Computer-aided assessment for enlarged fetal heart with deep learning model
Siti Nurmaini1, Ade Iriani Sapitri1, Muhammad Taufik Roseno2
1Intelligent System Research Group, Universitas Sriwijaya, Palembang, Indonesia.
Iscience
|May 9, 2025
Summary
This study introduces an automated deep learning tool for detecting fetal heart enlargement using YOLOv8. This AI approach enhances accuracy and consistency in prenatal ultrasound screenings, aiding early diagnosis of congenital heart conditions.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Cardiology
Background:
- Enlarged fetal heart conditions can signal serious congenital heart diseases or other complications.
- Early detection via prenatal ultrasound is crucial but current manual assessments are subjective and inconsistent.
- Automating this assessment can improve diagnostic accuracy and efficiency.
Purpose of the Study:
- To develop and evaluate a deep learning model for automated fetal heart enlargement assessment.
- To compare the performance of different YOLO architectures and enhancements for this task.
- To provide a reliable computer-aided tool for sonographers in prenatal screenings.
Main Methods:
- Utilized the You Only Look Once (YOLO) deep learning architecture, specifically YOLOv8.
- Incorporated a Convolutional Block Attention Module (CBAM) and ResNeXtBlock for performance enhancement.
- Trained and tested the model on a dataset of fetal ultrasound videos.
Main Results:
- YOLOv8 with CBAM outperformed YOLOv11 with self-attention in fetal heart enlargement detection.
- The addition of ResNeXtBlock further improved model accuracy and prediction consistency.
- The developed model demonstrated strong capabilities in identifying fetal heart enlargement.
Conclusions:
- The proposed deep learning approach offers a reliable, automated method for fetal heart enlargement assessment.
- This AI tool has the potential to enhance prenatal care by enabling earlier and more accurate diagnoses.
- Further validation is needed to confirm clinical applicability and improve neonatal health outcomes.
