Evaluating CESTR* depends on evaluating the model at a particular frequency which we cannot be sure is in the measured z-spectrum (although it usually is).
We have so far achieved this by interpolating on the values in the z-spectrum. There was a bug in the code that failed to take account of the fact that the measured spectrum might not be ordered. This is now fixed but another issue remains, namely that there may be multiple B1 values in the spectrum which will then get mixed together when we sort the spectrum and used interchangeably in the interpolation - even though the z-spectrum is slightly different at different B1 values.
What would be better is to evaluate the model directly with a specified frequency. This is complicated by the fact that the wvec is a member variable. So we either have to factor out the calculation so this can be passed as a parameter or (easier) just save it, change it for the CESTR* calculation, and then change it back afterwards.
Would still need to decide what B1 power to use in this calculation if there are multiple ones in the input dataspec. We would basically have to choose one and then issue a warning saying which we chose.
Evaluating CESTR* depends on evaluating the model at a particular frequency which we cannot be sure is in the measured z-spectrum (although it usually is).
We have so far achieved this by interpolating on the values in the z-spectrum. There was a bug in the code that failed to take account of the fact that the measured spectrum might not be ordered. This is now fixed but another issue remains, namely that there may be multiple B1 values in the spectrum which will then get mixed together when we sort the spectrum and used interchangeably in the interpolation - even though the z-spectrum is slightly different at different B1 values.
What would be better is to evaluate the model directly with a specified frequency. This is complicated by the fact that the wvec is a member variable. So we either have to factor out the calculation so this can be passed as a parameter or (easier) just save it, change it for the CESTR* calculation, and then change it back afterwards.
Would still need to decide what B1 power to use in this calculation if there are multiple ones in the input dataspec. We would basically have to choose one and then issue a warning saying which we chose.