The METU-FP Toolkit

Introduction

The METU-FP Toolkit is a collection of programs for solving the Forward Problem (FP) of the Electro-Magnetic Source Imaging of the human brain using realistic head models. The FP of EMSI is the solution of electrical potential on the surface of the head, and magnetic fields outside the head for a given electrical (dipole) source inside the head. The FP solutions are used in the inverse problem (IP) to search for sources that match a given set of measurements. An accurate FP solution is one of the requirements for accurate source imaging.
The toolkit contains implementations of two numerical methods, the Finite Element Method (FEM) and the Boundary Element Method (BEM). Which are the two most widely used numerical technniques for EMSI forward problem. Each of these numerical methods receive a mesh representing of the geometry and electrical properties of the human head as input, and produce a coefficient matrix from the mesh definition. By solving the matrix equation Ax=b using the coefficient matrix A for a right-hand-side vector A representing the source term, it is possible to compute the unknown potential values X corresponding to the nodes of the mesh.
The FEM models complete head volume using volume elements. Each element is assigned tissue properties, such as conductivity at this region in space. The FEM model has a very large number of elements, but the coefficient matrix is sparse.
The BEM models only the boundaries between various tissues in the head, using triangular surface elements. The regions separated by the boundaries are considered to have uniform conductivity. The BEM mesh, and the resulting BEM coefficient matrix has much less number of nodes. However the coefficient matrix is dense.

Documentation

The Toolkit consists of two parts, the BEM solver is called METU-BEM. The FEM solver is called METU-FEM.

METU-BEM

The BEM implementation uses linear and quadratic triangular surface elements. It supports the following advanced features:

The ability to support intersecting tissue boundaries,
Recursive integration for improved accuracy,
Linear and quadratic (curved) elements,
Isolated Problem (IPA) to reduce the numerical errors caused by the low conductivity of the skull layer

Mesh Format

The surface mesh used by METU-BEM consists of a set of three ASCII text files.

.bei Information file
.bec Coordinate file
.bee Connectivity (element) file

Matrix Creation: bem_matrix

The METU-BEM toolkit contains a utility called bem_matrix for creating matrices from mesh files and conductivity definitions. This utility is suitable for generating and saving matrices for use with an external solver, such as MATLAB.

Usage: bem_matrix [-f matname] [-m magsens] [-o mod] 
               meshname [s=sig ...]
       -f: save matrices to matname.[ci]mat
	    (default meshname)
       -m: magnetic sensor file name
       -o: interface to use with modified equations
           (1: outer, 0: to disable)
	s=sig:	s:region (1: outer), sig: conductivity
           (default conductivity: 0.2 for all regions)

Matrix Solution: bem_solve

Another, utility bem_solve, can be used to generate FP solutions for a given set of dipoles. It can also be used to generate a transfer matrix for a given set of sensor nodes (electrodes).

Usage: bem_solve [-m] [-l load_file] [-s save_file] [-o mod] 
               [-p potential_file] [-r sensor_file]
               [-d dipole_file] meshname [s=sig ...]

       -l: load matrices from matname.[ci]mat
       -s: save matrices to matname.[ci]mat
       -m: load Cmat only from matname.cmat
       -d: load dipoles from file
       -p: potential filename (base)
       -r: list of nodes to invert for reciprocity
           saved to save_file.inv
       -o: interface to use with modified equations
           (-1 to disable)

METU-FEM

Under construction ...

Source Code

METU-BEM

metu_bem-0.1.tar.gz

METU-FEM

License

The METU-FP Toolkit is released under the terms of GPLv2 license.

Publications

METU-BEM

Z. Akalın-Acar, N. G. Gencer, “An advanced BEM implementation for the forward problem of Electro-Magnetic Source Imaging”, Physics in Medicine and Biology. Vol: 49, no: 5, pp: 5011-5028, 2004.
N. G. Gencer, Z. Akalın-Acar, “Use of the isolated problem approach for multi-compartment BEM models of electro-magnetic source imaging”, Physics in Medicine and Biology. Vol: 50, pp: 3007-3022, 2005.
Y Ataseven, Z. Akalin-Acar, C. E. Acar, N. G. Gencer, “Parallel implementation of the accelerated BEM approach for EMSI of the human brain”, Med. & Biol. Eng. & Comp., 2008.
Z. Akalın Acar, Electro-Magnetic Source Imaging Using Realistic Head Models, PhD Thesis, Department of Electrical and Electronics Engineering, METU, June 2005.

METU-FEM

Nevzat G. Gençer, Can E. Acar, Ihsan O. Tanzer, Forward Problem Solution of Magnetic Source Imaging, in Magnetic Source Imaging of the Human Brain, Lawrance Erlbaum Associates, Inc., Publishers, Ed: Zhong-Lin Lu, Lloyd Kaufmann, May 2003.
Nevzat G. Gençer, Can E. Acar, “Sensitivity of EEG and MEG measurements to tissue conductivity”, Phys. Med. Biol. 49, 701-717, 2004.