PHYCORR Pipeline

PHYCORR (Physiological Correction) contains a Matlab-based pipeline for performing R-DECO and RetroICOR analysis on physiological data. The pipeline consists of preprocessing steps to extract cardiac and respiratory signals using R-DECO, followed by the RetroICOR processing.

• The foundational code and main functions used in this pipeline were provided by the Slocco Laboratory at Spaulding Rehabilitation Hospital. https://www.martinos.org/investigator/roberta-sclocco/

Workflow

The physiological raw data were collected using four channels in LabChart: heartrate, respiration, stimulus triggers and MRI triggers.

The pipeline is executed in the following order:

1. Preprocessing:
   • preprocessing.m: Processes raw physiological data to prepare it for peak detection.
   • r-deco (included in repository): Used to extract physiological peaks (e.g., R-peaks). Specifically, execute R_DECO.m from the r_deco folder.
   • consolidation_GUI.m: Merges the preprocessed physiological data and the extracted peaks into a single .mat file.
   • generate_1D_main_GUI_v2.m: Generates 1D representations of the cardiac (QRS) and respiratory signals from the consolidated data. Requires generate_1D_fun_1.m to be in the Matlab path.
2. RetroICOR Processing:
   • retroicor.m: Performs the RetroICOR analysis using the generated 1D signals and BOLD fMRI data.

Detailed Instructions

1. Preprocessing

1. Run preprocessing.m:
   • Execute preprocessing.m in Matlab.
   • Input the raw physiological data.
2. Extract Peaks using r-deco:
   • Navigate to the r_deco folder within this repository.
   • Execute R_DECO.m in Matlab to identify and extract physiological peaks (e.g., R-peaks).
3. Run consolidation_GUI.m:
   • Execute consolidation_GUI.m in Matlab.
   • In the GUI:
     • Select the preprocessed physiological data.
     • Select the extracted peaks from r-deco.
     • Specify an output filename for the consolidated .mat file.
4. Run generate_1D_main_GUI_v2.m:
   • Ensure that generate_1D_fun_1.m is in the Matlab path (addpath('path/to/generate_1D_fun_1.m')).
   • Execute generate_1D_main_GUI_v2.m in Matlab.
   • In the GUI:
     • Select the folder containing the consolidated .mat files.
     • Specify the output file.
     • This step will generate 1D representations of the QRS and respiratory signals.

2. RetroICOR Processing

1. Add RetroICOR folder to path:
   • Add the folder containing retroicor.m to the Matlab path (addpath('path/to/retroicor')).
2. Run retroicor.m:
   • Execute retroicor.m in Matlab.
   • In the GUI:
     • Specify the folder containing the RetroICOR code.
     • Select the BOLD fMRI data in NIfTI format (from BIDS sourcedata).
     • Select the corresponding JSON file containing metadata for the BOLD data.
     • Select the 1D QRS signal generated by generate_1D_main_GUI_v2.m.
     • Select the 1D respiratory signal generated by generate_1D_main_GUI_v2.m.
     • Specify the output folder. Note: Due to a current issue, the output files may not be placed directly into this folder. You will need to manually move them.
3. Manual Output Handling:
   • After the RetroICOR processing, manually move the output files from the default location to the specified output folder.

Dependencies

• Matlab
• BIDS-compliant fMRI data
• Physiological data

Proprietary Mentions

• RetroICOR: This method is based on the original work by Glover et al. (2000), "Retrospective correction of physiological fluctuation in fMRI signals." NeuroImage 11, 162–175.
• R-DECO: This tool is based on the original work by Moeyersons, J. et al. (2020)," R-DECO: An open-source Matlab based graphical user interface for the detection and correction of R-peaks (version 1.0.0). PhysioNet. https://doi.org/10.13026/x6j7-sp58."
• This project builds upon base code and functions provided by the Slocco Laboratory at Spaulding Rehabilitation Hospital (Lizbeth Ayoub, Andrew Bolender, and Roberta Slocco).

Known Issues

• The retroicor.m script currently has a bug that prevents output files from being written directly to the specified output folder. This will be addressed in a future version.

Future Improvements

• Fix the output folder issue in retroicor.m.
• Improve error handling and input validation.
• Add more detailed documentation and examples.
• Automate the movement of the output files.