Related Experiment Video
Updated: May 28, 2026

Direct Linear Transformation for the Measurement of In-Situ Peripheral Nerve Strain During Stretching
Published on: January 12, 2024
Stretcher: a learning-based framework for deformation-robust keypoint descriptors
Constantin von Witzleben1,2, Nazim Haouchine3
1Brigham and Women's Hospital, Harvard Medical School, Boston, USA. constiwitzleben@gmail.com.
This study introduces Stretcher, a novel framework enhancing keypoint descriptors for robust surgical tracking and navigation. Stretcher improves accuracy in soft-tissue deformation, crucial for procedures like liver surgery.
Area of Science:
- Medical image analysis
- Computer vision in surgery
- Surgical navigation systems
Background:
- Keypoint-based tracking is vital for image-guided surgery.
- Soft-tissue deformation, common in liver surgery, degrades tracking reliability.
- Existing keypoint descriptors struggle with large affine and non-rigid transformations.
Purpose of the Study:
- To introduce Stretcher, a framework enhancing deformation robustness of keypoint descriptors.
- To improve reliability of surgical tracking and navigation systems.
- To address limitations of current descriptors in handling tissue deformation.
Main Methods:
- Stretcher learns a neural model to capture the effect of affine deformations on descriptor representations.
- Keypoint descriptors are adjusted using the learned model, avoiding redundant recomputation.
- Simulates a grid of affine transformations for efficient descriptor adaptation.
Main Results:
- Stretcher improves keypoint matching robustness in highly deformed liver surgery images.
- Maintains state-of-the-art precision across evaluated scenarios.
- Demonstrates higher matching accuracy than existing methods in severe deformation.
Conclusions:
- Stretcher enables reliable keypoint matching under complex tissue motion by modeling affine deformations.
- The framework enhances robustness in challenging surgical scenarios without compromising efficiency.
- Well-suited for deformation-prone intraoperative tracking and navigation.
Related Concept Videos
Deformation of Member under Multiple Loadings
In the case of a member with a variable cross-section, the strain is not constant but depends on the position. The deformation of an...
Three-Dimensional Analysis of Strain
Transformation of Plane Strain
Under plane strain conditions, typical for members where one dimension significantly exceeds the others, deformations and resultant strains are...
Deformation of a Beam under Transverse Loading
The insights from the bending moment diagram extend to...
Measurements of Strain
¹³C NMR: Distortionless Enhancement by Polarization Transfer (DEPT)
