CRSNS: Development of EEG/MEG Source Reconstruction with Fast Multipole Method - This proposal entitled “Development of EEG/MEG Source Reconstruction with Fast Multipole Method” will
address the currently unmet need for significantly more accurate, near real-time EEG/MEG source
localization tools. The result will be the construction of an entirely new, high-speed, high-resolution
EEG/MEG source localization pipeline that includes: (i) field solution using the charge-based Boundary
Element Method (BEM) with Fast Multipole acceleration; (ii) inverse problem solution algorithm utilizing
the circuital reciprocity theorem and global cortical basis functions with an unlimited number of cortical
sources; (iii) precisely segmented head models and; (iv) precise MEG compatible EEG hardware,
verification and validation experiments and their documentation. If successful, the developed technology
will be adaptable to any branch of cognitive neuroscience diagnosis (EEG and MEG), providing enhanced
understanding of a patient’s response during antidepressant treatments and greater insight into overall
mental well-being.
Aim 1 Algorithm for EEG/MEG source reconstruction with fast multipole method
Aim 1A Algorithm construction and development (WPI Worcester and MPI Leipzig)
Aim 1B Preliminary results (WPI Worcester and MGH)
Aim 2 Integration with high density dry-electrode EEG system. EEG/MEG Measurements
Aim 2A Design a 256 channel MEG-compatible dry-electrode system (TU Ilmenau)
Aim 2B Perform MRI/EEG/MEG verification measurements (MPI Leipzig & TU Ilmenau)
Aim 2C Verify, modify and improve BEM-FMM based source reconstruction software (MPI Leipzig, WPI
Worcester, TU Ilmenau)
Coupling high-resolution algorithms with the finest yet robust and easy-to-use EEG hardware is one major
proposal goal. Such coupling could make an advanced high-density dry-electrode EEG very accurate.
