# EEG Sensor space metrics

**URL:** <https://mne.discourse.group/t/eeg-sensor-space-metrics/8333>\
**Category:** Support & Discussions\
**Tags:** source-localization, eeg\
**Created:** [February 14, 2024, 4:36pm UTC](https://mne.discourse.group/t/eeg-sensor-space-metrics/8333 "2024-02-14T16:36:25Z")\
**Posts on this page:** 7\
**Page:** 1

<div class="post-metadata">

**Author:** ![sarapiers](https://yyz2.discourse-cdn.com/free1/user_avatar/mne.discourse.group/sarapiers/32/3444_2.png) [@sarapiers](https://mne.discourse.group/u/sarapiers)\
**Post date:** [February 14, 2024, 4:36pm UTC](https://mne.discourse.group/t/eeg-sensor-space-metrics/8333/1 "2024-02-14T16:36:25Z")

</div>

Hello everyone,  
I am new to this so I hope I am not bothering you with my question.  
I am working with MNE Python stable version 1.6 (OS Windows10), to later use HNN core ([HNN Core | HNN](https://hnn.brown.edu/hnn-core/)), to fit the data and retrieve the underlying cellular and circuit-level neural generators.  
They provide one tutorial on how to fit real data ([04. From MEG sensor-space data to HNN simulation — hnn-core documentation](https://jonescompneurolab.github.io/hnn-core/stable/auto_examples/workflows/plot_simulate_somato.html#sphx-glr-auto-examples-workflows-plot-simulate-somato-py)) where the final label extracted time course results to be in (n)A\*m.  
I was trying to do the same with my EEG data but I do not understand what units of measurements I end up with in the source space - also how this would change among the different methods used (MNE, dSPM, etc).  
Can anyone help me understand this?  
I read the MNE documentation on “Inverse Problem and Source Analysis” but I did not find anything so far.

---

<div class="post-metadata">

**Author:** ![richard](https://yyz2.discourse-cdn.com/free1/user_avatar/mne.discourse.group/richard/32/15_2.png) [@richard](https://mne.discourse.group/u/richard)\
**Post date:** [February 14, 2024, 4:42pm UTC](https://mne.discourse.group/t/eeg-sensor-space-metrics/8333/2 "2024-02-14T16:42:51Z")

</div>

Hello and welcome to the forum!

Maybe this posting answers your question:

> [@What is the unit for the source estimates returned by apply\_inverse\_epochs function for EEG data?](https://mne.discourse.group/t/what-is-the-unit-for-the-source-estimates-returned-by-apply-inverse-epochs-function-for-eeg-data/3829/2):
>
> Hello @dahaiyu and welcome to the forum! The dSPM and sLORETA algorithms produce source estimates in unit-less values (based on F-statistics). On the other hand, the MNE and eLORETA algorithms return the electrical current in Amperes (that’s why the numbers are so tiny!). It is therefore totally alright to label the y-axis in your above figure as “dSPM value”. Hope that helps! Richard

Best wishes,  
Richard

---

<div class="post-metadata">

**Author:** ![sarapiers](https://yyz2.discourse-cdn.com/free1/user_avatar/mne.discourse.group/sarapiers/32/3444_2.png) [@sarapiers](https://mne.discourse.group/u/sarapiers)\
**Post date:** [February 14, 2024, 5:31pm UTC](https://mne.discourse.group/t/eeg-sensor-space-metrics/8333/3 "2024-02-14T17:31:13Z")

</div>

Thank you so much Richard.

And what about the " \*meter "? Is it just a different way to say the same thing?  
On the tutorial mentioned they use:

```
stc = apply_inverse(evoked, inv, lambda2, method='MNE',
                    pick_ori="normal", return_residual=False,
                    verbose=True)
flip_data = stc.extract_label_time_course(label_s1, inv['src'],
                                          mode='pca_flip')
dipole_tc = -flip_data[0] * 1e9

```

And refer to dipole\_tc as [nAm].  
Thank you again for your patience and for your time,  
Best,

Sara

---

<div class="post-metadata">

**Author:** ![richard](https://yyz2.discourse-cdn.com/free1/user_avatar/mne.discourse.group/richard/32/15_2.png) [@richard](https://mne.discourse.group/u/richard)\
**Post date:** [February 14, 2024, 9:10pm UTC](https://mne.discourse.group/t/eeg-sensor-space-metrics/8333/4 "2024-02-14T21:10:53Z")

</div>

Whoops, you might be right! And the response I cited above might be wrong. Indeed, the unit is Ampere\*meter. You can even check the unit of the inverse operator:

```python
# %%
import mne

sample_dir = mne.datasets.sample.data_path()
inv_op_fname = sample_dir / 'MEG' / 'sample' / 'sample_audvis-eeg-oct-6-eeg-inv.fif'
inv_op = mne.minimum_norm.read_inverse_operator(inv_op_fname)
print(inv_op["units"])

```

Produces:

```python
'Am'

```

Sorry about the confusion.

I’m not sure if the units eventually cancel out though? 😃 @larsoner or @agramfort, this is probably something you can answer from the top of your head 🙂

Best wishes,  
Richard

---

<div class="post-metadata">

**Author:** ![larsoner](https://yyz2.discourse-cdn.com/free1/user_avatar/mne.discourse.group/larsoner/32/3_2.png) [@larsoner](https://mne.discourse.group/u/larsoner)\
**Post date:** [February 16, 2024, 1:36pm UTC](https://mne.discourse.group/t/eeg-sensor-space-metrics/8333/5 "2024-02-16T13:36:54Z")

</div>

Yeah the units of MNE and eLORETA should be SI Ampere-meter. And once you take that and multiple by 1e9 nAm/Am the units are in nAm, so I think the tutorial is correct. This will be the same regardless of whether you use MEG, EEG, or M/EEG data.

---

<div class="post-metadata">

**Author:** ![sarapiers](https://yyz2.discourse-cdn.com/free1/user_avatar/mne.discourse.group/sarapiers/32/3444_2.png) [@sarapiers](https://mne.discourse.group/u/sarapiers)\
**Post date:** [February 16, 2024, 8:37pm UTC](https://mne.discourse.group/t/eeg-sensor-space-metrics/8333/6 "2024-02-16T20:37:47Z")

</div>

Thank you very much to everyone

---

<div class="post-metadata">

**Author:** ![system](https://global.discourse-cdn.com/free1/uploads/mne/original/1X/85cc6bd2b69cb698a166dc6d880fb550510d0144.jpeg) [@system](https://mne.discourse.group/u/system)\
**Post date:** [February 23, 2024, 8:38pm UTC](https://mne.discourse.group/t/eeg-sensor-space-metrics/8333/7 "2024-02-23T20:38:30Z")

</div>

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.