Support for 3-wavelength SNIRF files (e.g., HuiChuang fNIRS) in mne-nirs

[Liam6130]

Dear mne-nirs developers,

First, thank you for your excellent work on expanding MNE’s support for fNIRS analysis. The toolbox has been extremely helpful for our research workflow.

I would like to report a limitation in the current read_raw_snirf implementation. At present, mne-nirs appears to assume two wavelengths (the typical dual-wavelength CW-fNIRS setup), and the SNIRF reader cannot correctly load data from devices that use three wavelengths.

However, several manufacturers—such as HuiChuang Medical (慧创医疗) in China—have already released 3-wavelength fNIRS systems. These files follow the SNIRF standard but include three distinct wavelengths (e.g., 730/808/850 nm). When we load these SNIRF files in mne-nirs, the software either fails to parse the data or misinterprets the channel structure, making downstream HbO/HbR processing impossible.

Request

Could you consider updating mne-nirs to support SNIRF files containing three or more wavelengths?

This would greatly benefit researchers who rely on multi-wavelength systems for more accurate chromophore separation or additional biomarkers.

Additional information
• Device example: HuiChuang NirSmart (3-wavelength CW-fNIRS, SNIRF output)
• Wavelengths: 730 / 808 / 850 nm
• I can provide a minimal SNIRF file for testing if needed.

Thank you again for your outstanding contribution to the fNIRS community!
Best Regards
Liam

[larsoner]

Would mne-tools/mne-python#13408 help when it lands?

[zEdS15B3GCwq]

FYI, the PR mentioned above has been merged so if you install mne-python's main branch from GitHub, loading 3-wavelength files should work. As the author of that PR I'd appreciate any feedback on how the new code works with data from devices other than the one I tested on (Shimadzu Labnirs). If you can upload a small test file, that's also great. I hope to start working on similar changes in mne-nirs soon.