ROAST: An Open-Source, Fully-Automated, Realistic Volumetric-Approach-Based Simulator For TES

Area of Science:

• Neuroscience
• Computational Modeling
• Biomedical Engineering

Background:

• Transcranial electrical stimulation (TES) research uses computational models for current flow analysis.
• These models require detailed individual head anatomy from MRI scans, involving segmentation and electrode placement.
• Existing software pipelines offer partial solutions for these complex modeling steps.

Purpose of the Study:

• To introduce ROAST, an open-source, end-to-end pipeline for automated head modeling in TES.
• To leverage open-source tools (SPM8, iso2mesh, getDP) and custom scripts for improved segmentation and electrode placement.
• To validate ROAST's accuracy against established commercial finite element modeling software.

Main Methods:

• Development of ROAST, a fully automated processing pipeline for individual head models.
• Utilizing magnetic resonance images (MRIs) for realistic volumetric anatomy.
• Integration of SPM8, iso2mesh, and getDP, with custom scripts for segmentation and electrode placement.
• Finite element modeling to simulate and estimate electric current flow.

Main Results:

• ROAST provides an automated pipeline for processing individual head MRIs for TES modeling.
• Comparison with commercial software (ScanIP, Abaqus) showed a small 9% difference in electric field estimation.
• A 47% difference was observed when compared to SimNIBS, which uses surface segmentation.

Conclusions:

• ROAST offers an accurate and automated solution for computational modeling of TES.
• The pipeline's open-source nature facilitates wider adoption and research in TES.
• ROAST demonstrates reliable performance, comparable to commercial finite element modeling tools.