Related Experiment Video
Updated: Aug 28, 2025
06:56
Human Fetal Blood Flow Quantification with Magnetic Resonance Imaging and Motion Compensation
Published on: January 7, 2021
2.5K
Ensemble Transfer Learning for Fetal Head Analysis: From Segmentation to Gestational Age and Weight Prediction
Mahmood Alzubaidi1, Marco Agus1, Uzair Shah1
1College of Science and Engineering, Hamad Bin Khalifa University, Doha P.O. Box 34110 , Qatar.
Diagnostics (Basel, Switzerland)
|September 23, 2022
Summary
This study introduces an AI framework for analyzing fetal ultrasound images to accurately estimate gestational age and fetal weight. The model achieves high accuracy in fetal head segmentation and measurements, improving prenatal care diagnostics.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Obstetrics
Background:
- Ultrasound is crucial in obstetrics for monitoring fetal growth and detecting complications.
- Automated analysis of medical images using AI and computer vision is advancing diagnostic capabilities.
Purpose of the Study:
- To develop an end-to-end AI framework for segmenting, measuring, and estimating fetal gestational age and weight from 2D ultrasound images.
- To enhance the accuracy and efficiency of prenatal diagnostics through automated analysis of fetal head ultrasound data.
Main Methods:
- A novel framework combining eight segmentation architectures fine-tuned with EfficientNetB0 and an optimized ensemble transfer learning model (ETLM).
- ETLM was employed for fetal head segmentation and precise measurements, integrated into a multiple regression model for gestational age and estimated fetal weight (EFW) prediction.
- Validation involved comparison with expert physicians and longitudinal references using the HC18 dataset.
Main Results:
- Achieved 98.53% mean intersection over union (mIoU) for fetal head segmentation, surpassing state-of-the-art methods.
- Demonstrated measurement accuracy with a 1.87 mm mean absolute difference (MAD).
- Obtained low prediction errors: 0.03% MSE for gestational age and 0.05% MSE for EFW.
Conclusions:
- The developed AI framework provides highly accurate segmentation and measurement of fetal head ultrasound images.
- This automated approach significantly improves the prediction of gestational age and fetal weight, offering a valuable tool for prenatal care.
