logpdf_mcDESPOT_NoEx.m

%%% Function for search space sampling. Assumes on-resonance. %%%

% Inputs:
%         x: Parameter values at which to sample cost-function.
%         Lower: Lower bounds for search space sampling.
%         Upper: Upper bounds for search space sampling.
%         FA_SPGR: Array of flip angles for SPGR (radians).
%         FA_SSFP0: Array of flip angles for bSSFP0 (radians).
%         FA_SSFP180: Array of flip angles for bSSFP180 (radians).
%         TR_SPGR: Repetition time for SPGR (s).
%         TR_SSFP: Repetition time for bSSFP (s).
%         Data: Ground-truth simulated data, concatenated as [SPGR ; SSFP0 ; SSFP180].
%         Sigma: Notional noise variance.

% Outputs:
%         lnP: Value of pseudo-likelihood function.


function lnP = logpdf_mcDESPOT_NoEx(x, Lower, Upper, FA_SPGR, FA_SSFP0, FA_SSFP180, TR_SPGR, TR_SSFP, Data, Sigma)

if any(x < Lower) || any(x > Upper)
  lnP = -inf;
  return
end

[Mss_SPGR, Mss_SSFP0, Mss_SSFP180] = MEX_mcDESPOT_B0_DE_DiffFAs(x(2), x(1), x(5), x(4), 0, x(3), TR_SPGR, TR_SSFP, 0, rad2deg(FA_SPGR), rad2deg(FA_SSFP0), rad2deg(FA_SSFP180));

Mss_SSFP = [Mss_SSFP0 , Mss_SSFP180];

Data_SPGR = Data(1:numel(Mss_SPGR));
Data_SSFP = Data(numel(Mss_SPGR)+1:end);

Data_SPGR = (Data_SPGR * mean(Mss_SPGR))./mean(Data_SPGR);
Data_SSFP = (Data_SSFP * mean(Mss_SSFP))./mean(Data_SSFP);

Sig = [Mss_SPGR, Mss_SSFP]';
Data = [Data_SPGR; Data_SSFP];

lnP = -(sum((Sig - Data).^2))./(2*Sigma^2);

end