Semi-automatic 10/20 Identification Method for MRI-Free Probe Placement in Transcranial Brain Mapping Techniques

Xiang Xiao, Hao Zhu, Weijie Liu, Xiaoting Yu, Lian Duan, Zheng Li, Chaozhe ZhuFrontiers in Neuroscience 2017 · DOI

Abstract

The International 10/20 system is an important head-surface-based positioning system for transcranial brain mapping techniques, e.g., fNIRS and TMS. As guidance for probe placement, the 10/20 system permits both proper ROI coverage and spatial consistency among multiple subjects and experiments in a MRI-free context. However, the traditional manual approach to the identification of 10/20 landmarks faces problems in reliability and time cost. In this study, we propose a semi-automatic method to address these problems. First, a novel head surface reconstruction algorithm reconstructs head geometry from a set of points uniformly and sparsely sampled on the subject's head. Second, virtual 10/20 landmarks are determined on the reconstructed head surface in computational space. Finally, a visually-guided real-time navigation system guides the experimenter to each of the identified 10/20 landmarks on the physical head of the subject. Compared with the traditional manual approach, our proposed method provides a significant improvement both in reliability and time cost and thus could contribute to improving both the effectiveness and efficiency of 10/20-guided MRI-free probe placement.

Introduction

In transcranial brain mapping—techniques like functional Near-Infrared Spectroscopy (fNIRS) or Transcranial Magnetic Stimulation (TMS)—precision is paramount. To ensure researchers are stimulating or recording from the correct cortical regions, they rely on the International 10/20 system, a standardized coordinate system for the scalp.

While MRI-based localization is the gold standard, it is often inaccessible due to cost and time constraints. The common alternative is manual measurement using a tape measure, which is slow (averaging 16 minutes), prone to human error, and requires significant expertise. This study introduces a semi-automatic system to replace these tedious manual measurements with a faster, more reliable digital workflow.

Methodology: The S3R_VM Workflow

The proposed method, termed S3R_VM (Sparse Sample Surface Reconstruction-based Virtual Measurement), consists of three integrated steps:

1. Sparse Sampling

Using a magnetic 3D digitizer, an experimenter "clicks" approximately 21 points uniformly across the subject's head. To simplify this process, a standard EEG cap can serve as a guide for these initial points.

Sparse sampling of 3D points on the head surface.

2. Head Surface Reconstruction (S3R)

The core innovation is the reconstruction algorithm. Standard "crust algorithms" often fail to produce smooth surfaces from sparse data. By incorporating a spherization process—assuming the human head is roughly ellipsoidal—and non-linear interpolation, the system generates a high-density 3D model from the sparse input points.

Reconstructed head surface mesh generated from sparse points.

3. Virtual Measurement and Navigation

Once the digital model is created, 10/20 landmarks are automatically calculated in virtual space. A real-time navigation system then guides the experimenter to mark these points on the subject’s physical head, functioning essentially as a "GPS for the scalp."

Experimental Validation

The system was validated using:

  • MNI_ICBM_152 Head Model: Used to determine that 21 points provide the optimal balance between sampling effort and accuracy.
  • Real MR Images: 23 anatomical datasets provided "ground truth" geometries for simulation.
  • Physical Head Model: A medical-grade model was used to compare the performance of four different human operators against traditional manual methods.

Key Results

  • High Accuracy: The system achieved a mean estimation error of less than 3 mm with only 21 sample points, well within the requirements for fNIRS (30 mm) and TMS (10–20 mm).
  • Enhanced Reliability: Total variance was reduced from 6.41 mm (manual) to 2.93 mm (navigation), a 50% improvement in consistency.
  • Increased Efficiency: After a 1.5-minute initial sampling, navigating to any specific landmark takes only 28 seconds.

Discussion and Limitations

A key advantage of the S3R method is its robustness against operator inexperience. Unlike manual measurement, which is sequential (where one error cascades into all subsequent points), the S3R method calculates all points simultaneously.

However, the method requires specialized hardware (a 3D digitizer). Additionally, while the system automates much of the process, it still relies on the manual identification of four initial reference points (nasion, inion, and preauricular points). Future iterations could focus on automating the detection of these anatomical landmarks to further reduce human error.

Conclusion

The semi-automatic S3R_VM system offers a significant leap forward for MRI-free brain mapping. By halving the margin of error and dramatically speeding up the setup process, it provides a practical solution for high-precision probe placement in clinical and research settings.

Summary updated

Abstract

The International 10/20 system is an important head-surface-based positioning system for transcranial brain mapping techniques, e.g., fNIRS and TMS. As guidance for probe placement, the 10/20 system permits both proper ROI coverage and spatial consistency among multiple subjects and experiments in a MRI-free context. However, the traditional manual approach to the identification of 10/20 landmarks faces problems in reliability and time cost. In this study, we propose a semi-automatic method to address these problems. First, a novel head surface reconstruction algorithm reconstructs head geometry from a set of points uniformly and sparsely sampled on the subject's head. Second, virtual 10/20 landmarks are determined on the reconstructed head surface in computational space. Finally, a visually-guided real-time navigation system guides the experimenter to each of the identified 10/20 landmarks on the physical head of the subject. Compared with the traditional manual approach, our proposed method provides a significant improvement both in reliability and time cost and thus could contribute to improving both the effectiveness and efficiency of 10/20-guided MRI-free probe placement.
BibTeX
@inproceedings{xiao2017semi,
  author = {Xiang Xiao and Hao Zhu and Weijie Liu and Xiaoting Yu and Lian Duan and Zheng Li and Chaozhe Zhu},
  title = {{Semi-automatic 10/20 Identification Method for MRI-Free Probe Placement in Transcranial Brain Mapping Techniques}},
  booktitle = {Frontiers in Neuroscience},
  year = {2017},
  doi = {10.3389/fnins.2017.00004},
}
Figure