Related Experiment Video
Updated: Jul 31, 2025
09:32
Cortical Source Analysis of High-Density EEG Recordings in Children
Published on: June 30, 2014
21.4K
Harmonization of Multi-site Cortical Data Across the Human Lifespan
Sahar Ahmad1, Fang Nan2, Ye Wu1
1Department of Radiology and Biomedical Research Imaging Center (BRIC), The University of North Carolina at Chapel Hill, NC, USA.
Machine Learning in Medical Imaging. MLMI (Workshop)
|May 1, 2023
Summary
A new statistical method harmonizes multi-site neuroimaging data across the lifespan. This technique removes non-biological variations in cortical thickness and myelination, enabling robust developmental and aging research.
Area of Science:
- Neuroscience
- Biostatistics
- Medical Imaging Analysis
Background:
- Neuroimaging data harmonization is crucial for multi-site studies.
- Inconsistent data acquisition and processing introduce non-biological variability.
- This complicates the analysis of developmental and aging processes.
Purpose of the Study:
- To develop a novel statistical technique for retrospective harmonization of multi-site cortical data.
- To address data collected longitudinally and cross-sectionally from birth to 100 years.
- To eliminate non-biological disparities while preserving biological variation.
Main Methods:
- A novel statistical harmonization technique was developed.
- The method was applied to multi-site cortical thickness and myelination data.
- Longitudinal and cross-sectional data spanning birth to 100 years were analyzed.
Main Results:
- The proposed method effectively eliminated non-biological disparities in cortical measurements.
- Biological variation across the lifespan was successfully preserved.
- Harmonized data demonstrated improved consistency for statistical analyses.
Conclusions:
- The novel statistical technique provides effective retrospective harmonization of multi-site cortical neuroimaging data.
- This method enables the investigation of developmental and aging processes across the lifespan.
- The approach facilitates large-scale population studies by providing comparable data.
