Related Experiment Video
Updated: May 15, 2026
10:06
High-resolution Functional Magnetic Resonance Imaging Methods for Human Midbrain
Published on: May 10, 2012
A feature-based developmental model of the infant brain in structural MRI
Matthew Toews1, William M Wells, Lilla Zöllei
1Brigham and Women's Hospital, Harvard Medical School, USA. mt@bwh.harvard.edu
Medical Image Computing and Computer-Assisted Intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention
|January 5, 2013
Summary
This study models anatomical development using image patterns to predict infant age. The developed Bayesian model accurately identifies age-related brain structures and estimates age from MRI scans with a 72-day average error.
Area of Science:
- Medical Imaging
- Developmental Biology
- Computational Anatomy
Background:
- Accurate age estimation is crucial for tracking infant development.
- Understanding anatomical changes over time requires robust modeling techniques.
Purpose of the Study:
- To develop a computational model for anatomical development.
- To predict infant age using structural magnetic resonance imaging (MRI) data.
- To identify age-related anatomical structures.
Main Methods:
- Modeled anatomical development using spatio-temporal image patterns.
- Employed Bayesian posterior probability over subject age, conditioned on scale-invariant image features.
- Automatically learned the model from a large dataset of infant MRIs (230 scans, 92 subjects, 8-590 days).
Main Results:
- Successfully identified age-related anatomical structures in infant brains.
- Predicted the age of new subjects with an average error of 72 days.
- Demonstrated the model's ability to learn from diverse imaging data acquired at multiple sites.
Conclusions:
- The developed Bayesian model effectively captures anatomical development.
- This approach provides a reliable method for age prediction in infants using MRI.
- The findings contribute to a deeper understanding of early-life neurodevelopment.
