Add section on "Best Practices for Phys Data Collection" #177
Conversation
|
Hello there, any update on this? |
|
Ready for review! |
|
Great! Can I ask @RayStick, @BrightMG, and/or @CesarCaballeroGaudes to review this PR? |
|
I think that @BrightMG is looking into this with a PR to @AyyagariA's branch. |
Co-authored-by: Kristina Zvolanek <54590158+kristinazvolanek@users.noreply.github.com> Co-authored-by: Apoorva Ayyagari <50453337+AyyagariA@users.noreply.github.com>
|
@AyyagariA's PR now contains my updated version of the best practices doc. I'd appreciate another person reviewing this now |
Codecov Report
@@ Coverage Diff @@
## master #177 +/- ##
=======================================
Coverage 94.42% 94.42%
=======================================
Files 7 7
Lines 574 574
=======================================
Hits 542 542
Misses 32 32
|
|
I can have a read through soon, if that's welcome? |
|
Thanks Molly for your feedback and guidance! @kristinazvolanek and I will be adding references and making a few edits this week. |
I don't think I'll get to it today! @AyyagariA shall I wait until you've looked over it again, or shall I read when I can (probably tomorrow)? |
There aren't any major content changes for right now, so just read when you get a chance and we will let you know if there are updates! Thank you. |
docs/bestpractice.rst
Outdated
|
|
||
| **Breathing** is typically monitored using a "respiratory belt" around the participant's chest/diaphragm. The belt may be rigid or elastic, using MR compatible force or pressure transducers to generate a signal proportional to the chest diameter. The optimal positioning of the belt depends on the device being used, however it is best to be fairly consistent in how the belt is worn throughout a study. In some labs, multiple belts are used to better capture different types of breathing styles (e.g., "chest breathing" versus "belly breathing"). Often a belt is incorporated into the MRI scanner infrastructure, and these data can be collected by the scanner or recorded by a separate device. The peaks and troughs of the breathing trace are identified, which can provide information about breathing rate as well as breathing depth. There are three primary ways by which breathing can influence the fMRI signals. First, breathing often leads to bulk motion of the body and head. These effects are typically modeled using volume registration and motion correction algorithms. Second, breathing changes the chest position which can influence the success of the shim, continuously changing B0 homogeneity throughout the scan and in turn affecting signal amplitude. These effects are also modeled using techniques like RETROICOR. Thirdly, changing breathing rate and depth can influence blood gases, which can drive vasodilation or vasoconstriction, and thus substantially influence the fMRI signal amplitude. RVT correction (Birn et al. 2008) estimates the change in breathing rate/depth to model these effects. | ||
|
|
||
| **Blood gases:** It is also possible (and recommended!) to directly record changes in blood gas levels, rather than infer them from a chest position measurement. Most commonly we measure carbon dioxide levels (CO2), which is a known vasodilator and can drive large variability in blood flow and the BOLD signal. We can also measure oxygen (O2) levels; O2 only has a mild vasoconstictive effect on the cerebrovasculature, but oxygen levels can directly influence BOLD signal contrast. These two blood gases are typically strongly anticorrelated with eachother in most scans, but can also be manipulated independently and influence the fMRI signals through distinct mechanisms. Best practice would be to record both. Although the most accurate recordings of blood gas levels would be achieved through arterial sampling, this is not recommended for most imaging experiments. Instead, the concentrations of CO2 and O2 in arterial blood can be approximated by the partial pressure of each gas at the end of an exhalation, or the end-tidal partial pressure (commonly abbreviated as PETCO2 and PETO2). The person being scanned wears a nasal cannula (soft plastic tube that rests just below the nostrils) or face mask that is connected to a gas analyzer in the control room. The resulting data shows the fluctuations in CO2 and O2 across every breath; an algorithm must extract the "end-tidal" values. |
There was a problem hiding this comment.
It says 'oxygen' after introducing the abbreviation 'O2'
docs/bestpractice.rst
Outdated
| - respiratory belt | ||
| - disposable nasal cannula (or face mask) and long sample line to connect to control room | ||
|
|
||
| Some peripherals can be passed through a void in the penetration panel from the control room to the scan room (e.g., gas sampling line); others must be plugged into the penetration panel for noise filtering (e.g., some pulse sensors). Devices native to the MRI scanner may communicate wirelessly with the scanner. When adding non-native peripheral devices to the scanner environment, we recommend that you check that you are not bringing any outside noise into the scan room or bringing too much scanner noise into the physiological recordings. It may be necessary to develop additional devices or mechanisms to shield these connections. |
|
I gave it a read - reads well and good to go in my opinion (after references have been added, and formatting checked, etc.). There will always be more to say on this topic but I think this is a good concise introductory overview! There are a few very minor wording comments I made. |
|
@smoia This is ready to go! Just pushed the final touches. |
smoia
changed the title
|
@RayStick , if you approve, merge it in! |
Sure! (Was just reading the contributor file to remind myself whether I was allowed!) |
Closes #99 addresses #12
Proposed Changes