Related Experiment Videos
Curve matching on brain surfaces using Frenet distances
M Bakircioğlu1, U Grenander, N Khaneja
1Department of Electrical and Computer Engineering, Johns Hopkins University, Baltimore, Maryland 21218, USA.
Human Brain Mapping
|October 27, 1998
Summary
This study presents a novel method for matching curves on brain surfaces using diffeomorphic mapping and Frenet representation. The approach accurately aligns complex brain curves, aiding in comparative neuroanatomy research.
Area of Science:
- Neuroscience
- Computational Anatomy
- Medical Image Analysis
Background:
- Brain surface analysis requires robust methods for comparing anatomical features.
- Diffeomorphic matching offers a powerful framework for analyzing shape variations.
- Characterizing curves on brain surfaces is crucial for understanding cortical organization.
Purpose of the Study:
- To develop and demonstrate a method for diffeomorphic matching of curves on brain surfaces.
- To define and utilize Frenet representation for quantifying curve distances.
- To apply the method to fundus curves on macaque brain surfaces.
Main Methods:
- Utilized diffeomorphic mapping for curve alignment on brain surfaces.
- Defined curve distances using Frenet representation (speed, curvature, torsion).
- Employed bipartite graph matching with Frenet distances and dynamic programming for fundus curve identification.
Main Results:
- Successfully implemented a curve matching algorithm based on graph matching and Frenet distances.
- Identified and matched fundus curves, defined as extremal curvature lines.
- Demonstrated the method's application on macaque brain surface data.
Conclusions:
- The proposed method enables accurate diffeomorphic matching of curves on brain surfaces.
- Frenet representation provides a robust metric for curve comparison in neuroanatomy.
- This technique is valuable for comparative studies of brain structures and features.