Breathing Cycle and Brain

Looking for inspiration: new perspectives on respiratory rhythm. Jack L. Feldman and Christopher A. Del Negro (link)

Siemens Physiology - breathing movements; Chest sensor while person is in scanner supine.

Hi Jack,

I have a few questions that you may be able to help with, or point me in the right direction – I’ve done some searching without success.

Been working on a fine-grained look at fMRI in conjunction with a breathing signal (chest band). I’m interested in linking brain activity with phases of breathing – basically making a human version of Figure 1 in this paper of yours: NCBI - WWW Error Blocked Diagnostic . The question I have is how do I categorize the phases of the breathing cycle based on the human thoracic movement signal? In your paper you average ~100 respiratory cycles, but I’m not sure how to do this in our awake humans. I have the end-expiration/begin inspiration troughs and end inspiration/begin expiration peaks detected, so I can classify each timepoint as inspiration or expiration. Sample window for the fMRI is effectively 0.08 seconds (12.5Hz) so there is potential for separating out steps in the inspiration and expiration cycle. I’ve seen the code for the 2020 Lynch Nature Communications paper (https://www.nature.com/articles/s41467-020-18974-9 ). However, a couple of things, all related:

Different length breaths – is late expiration in a 6 second breath the same as late expiration in a 3 second breath (same person at rest)? I could classify expiration phase based on the time since begin-expiration, but should this be in seconds or as a percentage of the breath length? Should I restrict an analysis to a “normal” range of breath-lengths? (inter-quartile range maybe)
- Re person-to-person variation, I can normalize any absolute measures to a percentage of the individual’s median breath lengths throughout the recording
Peaks & troughs in the signal are quite flat – should these be separated out? (yellow boxes below) I’ve seen some people classify “end-inspiration” and “end-expiration” in addition to “inspiration” and “expiration”. On one level it makes sense to consider these times different, since intuitively these timepoints in flattish parts of the signal are equivalent to the mid-rise or mid-fall, which presumably would affect synchronized brain activity. However, their duration also seems to vary between people.
Different breath patterns – sighs and apneas should be separated, which is fine, but sometimes the shape of the breathing signal is quite different. Should there be some constraints on a “typical” breath…..? Could be positioning, could be individual differences in physiology or signal mechanics (e.g., the band is in a different shape when people of different sizes lie in the scanner).

Goal is to get this analysis running, and see what we can see in healthy people at rest, then hopefully get a paper together with everyone on board who would like to be. Then we can do measures wrt anxiety, breathing interventions acute and over time, etc. We could also do meditation, for the Feb ISCR symposium.

Below’s an example of what we get – there is a little pulse artifact in the breathing signal (“Resp”), and some scanner noise, but that’s rarely an issue.

A lot of questions I know, but any help would be useful!

Thanks,

Paul