HPI vs. cHPI - neuromag MEG system

Hello everyone,
I am using neuromag MEG data.
I have a .fif file I am working with.
If cHPI was used, does that mean that HPI was not used?
Another way to ask this: can HPI and cHPI be measured at the same scan?

Another question: - how do I reach the HPI data in MNE?

Thank you!

  • MNE version: 1.6.0
  • operating system: Windows 10

An HPI measurement is always performed at the beginning of the recording (although the operator can choose to omit the result. The HPI measurement will give you the initial device to head transformation.

cHPI will continuously send a waveform on the 5 HPI coils. The high-frequency content of those waveform is visible on the raw signal. To use the cHPI information, you can use MNE’s function to estimate in time:

(in this order, the output of the N-1 function is used in the N)

The head position information can then be used in MaxWell Filter (argument head_pos) to projects the signal onto the initial head position, i.e. the head position as measured by the initial HPI measurement.

The initial HPI measurement yields the device to head transformation stored in raw.info["dev_head_t"], and you can also find other HPI details in raw.info["hpi_meas "].



I really appreciate your response; it was very helpful.

Two questions, please:

  1. I want to make sure that the initial HPI did a good job. Is there a way to measure the accuracy of the calculation of the initial head positions?

  2. in raw.info["hpi_meas "] β†’ hpi_coils β†’ corr_coeff: what is the meaning of this correlation coefficient?

thank you very much!

I think you answered the first question with the second. Somewhere in the info structure, likely in raw.info["hpi_meas "] β†’ hpi_coils β†’ corr_coeff you should have the correlation coefficient between the estimated dipole position for each coil (as measured through the MEG sensors) and the digitized position of each coil (as measured by the polhemus).

It is unlikely that an operator started a recording with a badly fitted HPI measurement, the interface explicitly shows the goodness of fit and asks if we wish to accept this HPI measurement or do it again.


1 Like

thank you!