Related Experiment Video
Updated: Mar 23, 2026
High-resolution In Vivo Manual Segmentation Protocol for Human Hippocampal Subfields Using 3T Magnetic Resonance Imaging
Published on: November 10, 2015
Segmentation of Infant Hippocampus Using Common Feature Representations Learned for Multimodal Longitudinal Data
Yanrong Guo1, Guorong Wu1, Pew-Thian Yap1
1Department of Radiology and BRIC, University of North Carolina at Chapel Hill, NC, USA.
Insights
This study introduces a new method for segmenting the infant hippocampus using multimodal MRI scans. The approach improves accuracy in analyzing early brain development linked to conditions like ADHD.
Area of Science:
- Neuroimaging
- Developmental Neuroscience
- Medical Image Analysis
Background:
- Aberrant early brain development, particularly in the hippocampus, is linked to later-life neuropsychiatric disorders such as ADHD.
- Studying the infant hippocampus is challenging due to small size, variable image contrast, and inter-subject variability.
Purpose of the Study:
- To develop an effective method for hippocampus segmentation in infants using longitudinal multimodal MRI data.
- To address the challenges of analyzing heterogeneous infant brain imaging data.
Main Methods:
- A multi-atlas approach integrating longitudinal T1 and T2 MR images.
- Hierarchical multi-set kernel canonical correlation analysis (CCA) to learn common feature representations.
- Learning within-time-point and across-time-point common features for improved segmentation via label propagation and fusion.
Main Results:
- The proposed method demonstrates improved hippocampus segmentation performance compared to existing state-of-the-art techniques.
- Effective integration of complementary multimodal and longitudinal information was achieved.
Conclusions:
- The novel multi-atlas method offers enhanced hippocampus segmentation in infants.
- This technique has potential implications for understanding early brain development and associated disorders.
Abstract:
Aberrant development of the human brain during the first year after birth is known to cause critical implications in later stages of life. In particular, neuropsychiatric disorders, such as attention deficit hyperactivity disorder (ADHD), have been linked with abnormal early development of the hippocampus. Despite its known importance, studying the hippocampus in infant subjects is very challenging due to the significantly smaller brain size, dynamically varying image contrast, and large across-subject variation. In this paper, we present a novel method for effective hippocampus segmentation by using a multi-atlas approach that integrates the complementary multimodal information from longitudinal T1 and T2 MR images. In particular, considering the highly heterogeneous nature of the longitudinal data, we propose to learn their common feature representations by using hierarchical multi-set kernel canonical correlation analysis (CCA). Specifically, we will learn (1) within-time-point common features by projecting different modality features of each time point to its own modality-free common space, and (2) across-time-point common features by mapping all time-point-specific common features to a global common space for all time points. These final features are then employed in patch matching across different modalities and time points for hippocampus segmentation, via label propagation and fusion. Experimental results demonstrate the improved performance of our method over the state-of-the-art methods.
