Related Experiment Video
Updated: Sep 9, 2025
04:48
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
2.9K
Development of a Deep-Learning Model for Estimating Newborn Gestational Age via Lumbar Vertebral Segmentation on
Sungwon Ham1, Gayoung Choi2, Bo-Kyung Je3
1Healthcare Readiness Institute for Unified Korea, Korea University Ansan Hospital, Korea University College of Medicine, Seoul, Republic of Korea.
Korean Journal of Radiology
|August 28, 2025
Summary
This study developed a deep learning model to estimate newborn gestational age (GA) using lumbar vertebral body shapes from radiographs. The AI model shows potential for neonatal maturity assessment in clinical settings.
Area of Science:
- Medical Imaging
- Artificial Intelligence in Healthcare
- Neonatal Medicine
Background:
- Accurate estimation of gestational age (GA) is crucial for neonatal care.
- Traditional methods for GA assessment can be time-consuming or require specialized equipment.
Purpose of the Study:
- To develop and validate a deep learning model for estimating newborn GA.
- To assess the model's performance in segmenting lumbar vertebral bodies and estimating GA from radiographs.
Main Methods:
- Retrospective study of 423 neonatal radiographs.
- Development of a deep learning model using DeepLabv3+ and ResNet architectures for segmentation and GA estimation.
- External validation on a separate dataset of 167 radiographs.
Main Results:
- The segmentation model achieved a mean Dice similarity coefficient of 0.801 ± 0.031.
- The GA estimation model yielded a mean absolute error of 5.2 ± 0.5 days.
- Bland-Altman analysis showed a bias of -0.4 days with 95% limits of agreement from -13.0 to 12.3 days.
Conclusions:
- The deep learning model demonstrates promising performance for lumbar vertebral body segmentation and GA estimation.
- This AI tool has potential as a supportive method for assessing neonatal maturity in clinical practice.
