Influence of personalized human head modeling and resolution on EEG source localization for rapid brain mapping

Comprehensive Summary

This research examines the impact of electrode placement on image quality in electrical impedance tomography (EIT), which is a non-invasive imaging method that reconstructs the internal electrical conductivity from external measurements. To solve this problem, the authors proposed a deep learning-based approach to predict the electrode placement around a 2D object such that the voltage measurements obtained from such placement result in improved reconstructions. The authors tested their approach using simulated objects and compared the electrode placements obtained from their approach with standard placements. They found that the optimized placements resulted in improved reconstruction accuracy and better sketches of internal structures compared to standard placements. The authors also discussed the impact of optimized placements on sensitivity patterns and the robustness of the deep learning approach.

Outcomes and Implications

Optimizing the position of the electrodes can help make EIT a more reliable tool for various clinical applications, such as lung ventilation, edema detection, or tissue conductivity pattern changes. This is because EIT image quality has hindered the wider application of EIT in these areas. This work provides a way for improving the diagnostic capability of EIT without relying on hardware improvements by employing machine learning for smarter electrode configuration. Although this approach is promising, it still has to be verified in real-life biological scenarios. Its application in real-life devices may take a few years.