At every drift corr event we should update the frame such that the feature to be used for correlation is at the centre of the frame. These need to be saved along with the measured vectors for any post-mortem diagnosis.
Some other ideas:
- If each ROI only contains a NP, we could use the CoM of each data to be used to adjust the subsequent scans
- We can use the up-sampled survey image to drift correct with every dataset
At every drift corr event we should update the frame such that the feature to be used for correlation is at the centre of the frame. These need to be saved along with the measured vectors for any post-mortem diagnosis.
Some other ideas: