rshrf 1.6.3

BIDs App to retrieve the haemodynamic response function from resting state fMRI data

Resting state HRF estimation and deconvolution

Please refer to https://github.com/compneuro-da/rsHRF for MATLAB version

The basic idea

This toolbox is aimed to retrieve the onsets of pseudo-events triggering an hemodynamic response from resting state fMRI BOLD voxel-wise signal. It is based on point process theory, and fits a model to retrieve the optimal lag between the events and the HRF onset, as well as the HRF shape, using a choice of basis functions (the canonical shape with two derivatives, (smoothed) Finite Impulse Response, mixture of gammas).

Once that the HRF has been retrieved for each voxel, it can be deconvolved from the time series (for example to improve lag-based connectivity estimates), or one can map the shape parameters everywhere in the brain (including white matter), and use the shape as a pathophysiological indicator.

How to use the toolbox

The input is voxelwise BOLD signal, already preprocessed according to your favorite recipe. Important thing are:

• bandpass filter in the 0.01-0.08 Hz interval (or something like that)
• z-score the voxel BOLD time series

To be on the safe side, these steps are performed again in the code.

The input can be images (3D or 4D), or directly matrices of [observation x voxels].

It is possible to use a temporal mask to exclude some time points (for example after scrubbing).

The demos allow you to run the analyses on several formats of input data.

Python Package and BIDS-app

A BIDS-App has been made for easy and reproducible analysis. Its documentation can be accessed at:

https://bids-apps.neuroimaging.io/rsHRF/

Collaborators

• Guorong Wu

• Nigel Colenbier

• Sofie Van Den Bossche

• Daniele Marinazzo

• Madhur Tandon (Python - BIDS)

• Asier Erramuzpe (Python - BIDS)

• Amogh Johri (Python - BIDS)

Docker Usage

You can run rsHRF as a containerized BIDS App using Docker. This avoids the need to install Python dependencies locally and ensures reproducibility.

1. Pull the latest image:

docker pull bids/rshrf:latest

2. Run the analysis:

To run the analysis on a BIDS-formatted dataset, use the following command structure:

docker run -ti --rm \
  -v /path/to/your/bids_dataset:/data:ro \
  -v /path/to/your/output_dir:/out \
  bids/rshrf:latest \
  --bids_dir /data \
  --output_dir /out \
  --analysis_level participant \
  --participant_label 01 \
  --brainmask \
  --estimation canon2dd \
  --wiener

References

1. Wu, G. R., Colenbier, N., Van Den Bossche, S., Clauw, K., Johri, A., Tandon, M., & Marinazzo, D. (2021). rsHRF: A toolbox for resting-state HRF estimation and deconvolution. Neuroimage, 244, 118591. open access journal page

2. Guo-Rong Wu, Wei Liao, Sebastiano Stramaglia, Ju-Rong Ding, Huafu Chen, Daniele Marinazzo*. "A blind deconvolution approach to recover effective connectivity brain networks from resting state fMRI data." Medical Image Analysis, 2013, 17:365-374. Open access institutional repo

3. Guo-Rong Wu, Daniele Marinazzo. "Sensitivity of the resting state hemodynamic response function estimation to autonomic nervous system fluctuations." Philosophical Transactions of the Royal Society A, 2016, 374: 20150190. Open access institutional repo