This is a new chapter

This is a new chapter

Take home #0

This is a take home message

Cerebrovascular Reactivity (CVR)

Cerebrovascular Reactivity (CVR) is the response of cerebral vessels to a vasoactive stimulus (e.g. CO2) to provide sufficient O2 to cerebral tissues¹

1. Liu et al., 2018 (Neuroimage);    2. Pinto et al., 2021 (Front. Physiol.), Moia et al., 2021 (Neuroimage)

CVR can be measured during BOLD fMRI experiments with Breath-holds (BH), that induce the subject into a state of hypercapnia²

The vessels dilate → Increase of blood flow → Increase of %BOLD signal

Neurovascular Coupling

Image courtesy of Martin Havlicheck and Dimo Ivanov

Changes in
Haemodynamics

Neurovascular
Coupling

Changes in Oxygen Metabolism

Changes in
BOLD signal

(Venous) Vasculature

Arrows indicate causal influence

Recording physiological data

Ventilation

Respiration (O2/CO2)

Pulse

Task paradigm

Bright et al. 2013 (Neuroimage), Moia et al. 2021 (Neuroimage), Pinto et al. 2021 (Front. Physiol.)

Task paradigm

Bright et al. 2009 (Neuroimage), Pinto et al. 2021 (Front. Physiol.)

QA/QC

Image courtesy of Kristina Zvolanek

Subject compliance

1. Stickland et al., 2021 (NeuroImage);   2. Zvolanek et al., 2023 (Neuroimage)

  • Short respiratory challenges¹
  • Alternative estimates²

Easy additions!

Subject compliance

1. Stickland et al., 2021 (NeuroImage);   2. Zvolanek et al., 2023 (Neuroimage)

  • Short respiratory challenges¹
  • Alternative estimates²

Preprocessing

import os

import numpy as np
import peakdet as pk

[...]

data = np.genfromtxt(filename)
ph = pk.Physio(data[:, ch], fs=10000.0)

# Filter and downsample [optional]
ph = pk.operations.filter_physio(ph, 2, method='lowpass', order=7)
ph = pk.operations.interpolate_physio(ph, 40.0, kind='linear')

# Find peaks (adjust threshold and distance)
ph = pk.operations.peakfind_physio(ph, thresh=0.1, dist=60)

# Manually edit peaks - yes, you do need to at least check them.
ph = pk.operations.edit_physio(ph)

# Check the file history
ph.history

# Export the file
pk.save_physio(f"{os.path.splitext(filename)[0]}_co2.phys", ph)
  • Lowpass filter & downsample (opt)
  • Peak detection (and manual check!)
  • Export

CO2

  • No slice interpolation
  • No smoothing (optional)
  • No denoising
  • No filtering

fMRI

  • Segmentation (to get references)
  • Registration to fMRI

aMRI

Methods: Preprocessing

  1. MP2RAGE and T2w skullstripped (AFNI)
  2. MP2RAGE normalised to MNI152 template (ANTs)
  3. T2w registered to MP2RAGE and to SBREF (FSL)
  4. MP2RAGE segmented and GM registered to SBREF (ANTs)
  5. GM eroded (FSL)
  6. EPI realigned to SBREF, saved the motion parameters (ANTs)
  7. EPI skullstripped (FSL)
  8. EPI distortion corrections (FSL)

Make sure you're removing what you should

data = np.genfromtxt('sub-007_ses-05_task-rest_run-01_physio.tsv.gz', usecols=[0, 1, 3])
ph = peakdet.Physio(data[:, 1], fs=10000, suppdata=data[:, 2])
ph = peakdet.operations.peakfind_physio(ph, thresh=thr, dist=dist)
ph = peakdet.operations.edit_physio(ph)

DuPre et al., 2024 (Zenodo)

Ventilation

CO2

BH-induced CVR: Issues

Two problems:

  1. There's a measurement delay, and regional variations of the physiological delay
  2. The BH task presents collinear motion to signal of interest

Motion

CO2

BOLD

Moia et al. 2026 (Zenodo)

phys2cvr

  • CO2-based CVR estimation
  • (filtered) BOLD-based CVR estimation
  • Custom-based lag-GLM mapping
  • Interpolate end-tidal \(CO2\)
    \(CO_2 → P_{ET}CO_2\)
  • Convolve \(P_{ET}CO_2\) (H)RF
    \(P_{ET}CO_2hrf = P_{ET}CO_2 \ast hrf \)
  • First bulk-shift
    \(P_{ET}CO_2 \star BOLD_{GM}\)
  • Refined local shifts (lag GLM) with:
    • motion parameters
    • Legendre polynomials
    • other (orthogonalised) noise factors (e.g. conservative ICA denoising)

Supports

Lag

R²

CNR of lag maps SimMot SeqMot NoMot
GM-WM 0.52 ±0.21 0.46 ±0.26 0.49 ±0.25
GM-Putamen 0.47 ±0.22 0.44 ±0.21 0.44 ±0.21
GM-Cerebellum 0.82 ±0.15 0.69 ±0.17* 0.69 ±0.16*

Non optimising leads to underestimate the CVR, especially in subcortical areas.

Lag maps show anatomical consistency

→

→

→

→

→

→

→

→

Different lag responses, coherent with previous evidence¹ (e.g. Putamen has earlier response than GM)

→

→

→

Area mostly affected by motion!  →

Moia, Stickland, et al. 2020 (EMBC);    1. Blockley et al. 2011 (MRM), Bright et al. 2009 (Neuroimage)

Lag optimisation

phys2cvr tutorial

Parameters

1. Bright et al. 2013 (Neuroimage);    2. AFNI's afni_proc

Duration of apnea?

  • Can be reduced (down to 10s)¹ by increasing trials.
  • Better to make subjects comfortable and ask them to concede if unable to continue.

RF convolution?

  • May be skipped entirely, especially if hypocapnic transients are present in the response.

Degree of polynomials?

  • 1 + 1 every 150s of data.²
  • Use only if PETCO2 doesn't already capture slow variations.

Lags?

  • Can be increased to capture delayed hemodynamic responses.
  • Use caution if around or above half of the trial duration.

Stickland et al. 2021 (Neuroimage), Clemens et al. 2026 (bioRxiv)

Lag range

Moia, Stickland, et al. 2020 (EMBC)

Post-mapping thresholding

\alpha_{\text{sidak}} = 1 - (1 - \alpha)^{1/m}

Statistical threshold

Šidák correction for multiple comparisons

Lag-based threshold

Exclude non-physiological lag estimates

physiopy

Raw data

BIDSification

phys. data preprocessing
(peak detect.)

physiopy's documentation
&

phys. denoising

phys. imaging

BIDS Extension Proposal

Physiopy's Community Practices

QA/QC

github.com/physiopy
physiopy.github.io
physiopy-community-guidelines.rtfd.io
physiopy.community@gmail.com
s.moia.research@gmail.com
  • Toolboxes guarantee reproducibility even in presence of manual steps
  • Community guidelines guide new (and old) users in improving the quality and usage of physiological signals in neuroimaging

Any question [/opinions/objections/...]?

Stefano Moia, 2026

Find the presentation at:

slides.com/smoia/cvr_ll_2026/scroll