# AR-IRLS in mne-nirs

**URL:** <https://mne.discourse.group/t/ar-irls-in-mne-nirs/6990>\
**Category:** Support & Discussions\
**Tags:** fnirs\
**Created:** [May 31, 2023, 1:47pm UTC](https://mne.discourse.group/t/ar-irls-in-mne-nirs/6990 "2023-05-31T13:47:27Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![brizzen](https://yyz2.discourse-cdn.com/free1/user_avatar/mne.discourse.group/brizzen/32/2694_2.png) [@brizzen](https://mne.discourse.group/u/brizzen)\
**Post date:** [May 31, 2023, 1:47pm UTC](https://mne.discourse.group/t/ar-irls-in-mne-nirs/6990/1 "2023-05-31T13:47:27Z")

</div>

Is an implementation of AR-IRLS planned to mne-nirs based on Fig.2 from the Huppert’s, 2016 paper ([https://doi.org/10.1117/1.NPh.3.1.010401](https://doi.org/10.1117/1.NPh.3.1.010401)) comparing different models? It seems that AR(n) plus IRLS shows the best performance also in comparison to AR(1)-DCT.

Unlike in [https://mne.tools/mne-nirs/dev/auto\_examples/migration/plot\_02\_nirstoolbox.html](https://mne.tools/mne-nirs/dev/auto_examples/migration/plot_02_nirstoolbox.html), AR(n) was successfully implemented to run\_glm/nilearn ([https://github.com/mne-tools/mne-nirs/pull/277](https://github.com/mne-tools/mne-nirs/pull/277) - thanks for that:) ). However, I wonder whether IRLS provides any benefit in comparison to what is used right now?

- MNE version: 1.3.1
- mne-nirs: 0.5.0
- operating system: Windows 11

Best regards,  
Brais

---

<div class="post-metadata">

**Author:** ![sappelhoff](https://yyz2.discourse-cdn.com/free1/user_avatar/mne.discourse.group/sappelhoff/32/3276_2.png) [@sappelhoff](https://mne.discourse.group/u/sappelhoff)\
**Post date:** [June 13, 2023, 3:13pm UTC](https://mne.discourse.group/t/ar-irls-in-mne-nirs/6990/2 "2023-06-13T15:13:48Z")

</div>

I think this question should be a feature request on the mne-nirs GitHub repo: [GitHub - mne-tools/mne-nirs: Process Near-Infrared Spectroscopy Data in MNE](https://github.com/mne-tools/mne-nirs)