Home > functions > leadfield > eeg > vb_eeg_legendre.m

vb_eeg_legendre

PURPOSE ^

EEG potential for dipoles

SYNOPSIS ^

function eeg = vb_eeg_legendre(Xq, Q, Xe, R, sigma, Nmax)

DESCRIPTION ^

 EEG potential for dipoles 
     in multilayer spherical forward model 
 Legendre expansion calculation
  eeg = vb_eeg_legendre(Xq, Q, Xe, R, sigma, Nmax)
 INPUTS (Required):
   Xq   : dipole location(in meters)                  NP x 3
   Q    : dipole direction                            NP x 3
   Xe   : EEG sensors(in meters) on the scalp         M  x 3
   R    : radii(in meters) of sphere from 
          innermost to outermost                      NL x 1
   sigma: conductivity from innermost to outermost    NL x 1

 INPUTS (Optional):
     Nmax : # of terms used in Truncated Legendre Series

  M  = # of sensors
  NP = # of dipoles
  NL = # of sphere

 OUTPUTS:
   eeg : EEG Lead field matrix   ( NP x M )

 IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL.46, 245-259, 1999
 John C. Mosher, Richard M. Leahy and Paul S. Lewis

 Copyright (C) 2011, ATR All Rights Reserved.
 License : New BSD License(see VBMEG_LICENSE.txt)

CROSS-REFERENCE INFORMATION ^

This function calls: This function is called by:

SOURCE CODE ^

0001 function    eeg = vb_eeg_legendre(Xq, Q, Xe, R, sigma, Nmax)
0002 % EEG potential for dipoles
0003 %     in multilayer spherical forward model
0004 % Legendre expansion calculation
0005 %  eeg = vb_eeg_legendre(Xq, Q, Xe, R, sigma, Nmax)
0006 % INPUTS (Required):
0007 %   Xq   : dipole location(in meters)                  NP x 3
0008 %   Q    : dipole direction                            NP x 3
0009 %   Xe   : EEG sensors(in meters) on the scalp         M  x 3
0010 %   R    : radii(in meters) of sphere from
0011 %          innermost to outermost                      NL x 1
0012 %   sigma: conductivity from innermost to outermost    NL x 1
0013 %
0014 % INPUTS (Optional):
0015 %     Nmax : # of terms used in Truncated Legendre Series
0016 %
0017 %  M  = # of sensors
0018 %  NP = # of dipoles
0019 %  NL = # of sphere
0020 %
0021 % OUTPUTS:
0022 %   eeg : EEG Lead field matrix   ( NP x M )
0023 %
0024 % IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL.46, 245-259, 1999
0025 % John C. Mosher, Richard M. Leahy and Paul S. Lewis
0026 %
0027 % Copyright (C) 2011, ATR All Rights Reserved.
0028 % License : New BSD License(see VBMEG_LICENSE.txt)
0029 
0030 M  = size(Xe,1);    %  # of sensors
0031 NP = size(Xq,1);    %  # of dipoles
0032 NL = length(R);     %  # of sphere
0033 
0034 % Radius of dipoles
0035 Rq = sqrt(sum(Xq.^2,2));    % NP x 1
0036 
0037 % Radius of outermost layer (Sensor distance from origin)
0038 Rmax = R(NL);  
0039 
0040 if R(1)~= min(R)
0041     fprintf('Head radii must be R(1) < ... < R(NL)\n')
0042     return
0043 end
0044 
0045 if ~all( Rq < R(1) )    
0046     % check if dipoles within the  innermost layer
0047     fprintf('All dipoles must be inside innermost layer\n')
0048     return
0049 end
0050 
0051 % EEG potential
0052 eeg = zeros(NP,M);
0053 
0054 if nargin < 6, 
0055     % Default # of Legendre expansion
0056     Nmax = min( fix(10/log(Rmax/max(Rq))), 200);
0057 end
0058 
0059 % Normalized position (unit vector)
0060 Re = sqrt(sum(Xe.^2, 2));
0061 Xe = Xe./Re(:,ones(1,3));        % EEG sensor ( M x 3 )
0062 Xq = Xq./Rq(:,ones(1,3));        % dipole     ( NP x 3 )
0063 
0064 % Weight factor of legendre expansion
0065 f = vb_multi_shell_param(R, sigma, Nmax);    % Eq.(16) & (17)
0066 
0067 npow = 2*[1:Nmax]+1;
0068 
0069 % Coefficient of legendre expansion in Eq.(15)
0070 fn  = ( (npow./[1:Nmax]).*f' )/(4*pi*sigma(NL)*R(NL)^2); % ( 1 x Nmax )
0071 
0072 n   = [1:Nmax];
0073 nm1 = n-1;
0074 
0075 % loop for dipoles
0076 for i = 1:NP,  
0077     Xqn = Xq(i,:)';
0078     Qq  = Q(i,:);
0079     
0080     % angle between dipole moment and position vector
0081     cosalpha = Qq*Xqn;
0082     
0083     % angle between dipole position and sensor position vectors
0084     cosgamma = Xe*Xqn;    % ( M x 1 )
0085                 
0086     % evaluate legendre polynomial and its derivative
0087     [Pn, dP] = vb_legendre_grad(Nmax,cosgamma);  % ( Nmax x M )
0088     
0089     % power of radius
0090     ratio = (Rq(i)/Rmax).^nm1;
0091     
0092     z = Xe*Qq' - cosgamma*cosalpha;    % ( M x 1 )
0093     
0094     w = fn.*ratio;            %  (1 x Nmax)
0095     
0096     eeg1 = (w.*n) * Pn;        % (1 x M)
0097     eeg2 =  w * dP;            % (1 x M)
0098     
0099     eeg(i,:) = eeg1.*cosalpha  + eeg2.*z';
0100 end
0101 
0102 %eeg = eeg';

Generated on Tue 27-Aug-2013 11:46:04 by m2html © 2005