Related Experiment Video
Updated: Jul 18, 2026
09:30
Localizing Function-specific Targets for Transcranial Magnetic Stimulation in the Absence of Navigation Equipment
Published on: May 23, 2025
A cross validation study of deep brain stimulation targeting: from experts to atlas-based, segmentation-based and
F Javier Sanchez Castro1, Claudio Pollo, Reto Meuli
1Signal Processing Institute, Ecole Polytechnique Fédérale de Lausanne (EPFL), CH-1015 Lausanne, Switzerland. francisco.sanchezcastro@epfl.ch
IEEE Transactions on Medical Imaging
|November 23, 2006
Summary
Accurately locating the subthalamic nucleus (STN) for deep brain stimulation (DBS) is crucial for treating movement disorders. Our validation scheme shows automatic targeting is as accurate as expert methods, improving surgical precision.
Area of Science:
- Neurosurgery
- Medical Imaging
- Computational Anatomy
Background:
- Deep brain stimulation (DBS) is vital for movement disorder treatment.
- Accurate targeting of the subthalamic nucleus (STN) is essential but challenging due to poor visibility in standard medical images.
- Current methods rely on atlases and landmarks, requiring intraoperative adjustments.
Purpose of the Study:
- To develop and validate a novel scheme for evaluating image registration algorithms used in STN targeting for DBS.
- To compare the accuracy and usability of different atlas-based and non-rigid registration methods for automatic STN localization.
- To identify key anatomical structures influencing STN location estimation.
Main Methods:
- Constructed a ground truth dataset using MR T2-weighted images where STNs were clearly visible.
- Utilized a patient-specific atlas approach for registering other patients.
- Evaluated multiple registration algorithms and analyzed the influence of surrounding structures on STN localization accuracy.
- Cross-validated results against expert variability.
Main Results:
- Demonstrated that automatic STN targeting is feasible and achieves accuracy comparable to expert-driven techniques.
- Quantified the performance and usability of various registration algorithms for STN localization.
- Identified specific visible anatomical structures crucial for accurate STN estimation.
Conclusions:
- The proposed validation scheme effectively assesses image registration algorithms for DBS targeting.
- Automatic STN targeting is a viable and accurate alternative to current expert-based methods.
- Understanding the influence of surrounding anatomical landmarks is key to improving targeting precision.
