Related Experiment Video
Updated: Jan 11, 2026
05:33
Three-Dimensional Shape Modeling and Analysis of Brain Structures
Published on: November 14, 2019
7.6K
Digital tools for analyzing nondiffeomorphic shapes
Henry Kirveslahti1,2,3, Xiaohan Wang1,4
1Department of Mathematics and Computer Science, Laboratory for Topology and Neurosciences, Ecole polytechnique fédérale de Lausanne, Lausanne CH-1015, Switzerland.
Summary
We developed a precise digital algorithm for the Euler Characteristic Transform (ECT), enabling exact shape analysis without information loss. This new method enhances shape inversion and alignment for complex datasets.
Area of Science:
- Computational geometry
- Shape analysis
- Digital topology
Background:
- The Euler Characteristic Transform (ECT) is a landmark-free method for analyzing nondiffeomorphic shapes.
- Current discrete approximations of ECT lead to information loss and inversion challenges.
Purpose of the Study:
- To present a fully digital algorithm for exact ECT computation.
- To introduce the Ectoplasm package for implementing the algorithm.
- To demonstrate the algorithm's utility in real-world shape analysis tasks.
Main Methods:
- Developed a novel, fully digital algorithm for exact Euler Characteristic Transform computation.
- Implemented the algorithm in the Ectoplasm software package.
- Applied gradient descent and adaptive grid search for shape alignment using the exact ECT.
Main Results:
- The algorithm computes ECT exactly, up to computer precision.
- The Ectoplasm package offers a fast and convenient tool for computing distances on real-life datasets.
- Enabled shape inversion, subshape selection, and shape alignment previously not possible with discretized ECT.
Conclusions:
- The exact digital ECT algorithm overcomes limitations of previous approximations.
- The Ectoplasm package facilitates advanced shape analysis, including alignment.
- This work opens new possibilities for statistical shape analysis and related problems.
