Perfusion (ASL) Pipeline¶

The BRC Perfusion Pipeline (BRC_perfusion_pipeline) processes single-TI pseudo-continuous arterial spin labelling (pCASL) data to produce quantified cerebral blood flow (CBF) maps in both native and standard space, with optional partial volume correction.

The pipeline takes a difference (perfusion-weighted) image and a calibration (M0) image as inputs. It performs CBF quantification using a standard kinetic model, registers the ASL data to the subject's T1 anatomical image (using boundary-based registration leveraging the white matter segmentation), propagates the result to MNI152 standard space, organises outputs, and optionally applies multi-tissue-linear-series (MLTS) partial volume correction.

Quick Start¶

export BRC_GLOBAL_SCR=/opt/brc/global/scripts
export BRC_PMRI_SCR=/opt/brc/BRC_perfusion_pipeline/scripts
export FSLDIR=/usr/local/fsl
source ${FSLDIR}/etc/fslconf/fsl.sh

ASL_preproc.sh \
    --input_pwi  /data/sub-001/perf/sub-001_pwi.nii.gz \
    --input_m0   /data/sub-001/perf/sub-001_m0.nii.gz \
    --path       /data/study \
    --subject    sub-001 \
    --tr 3.2 \
    --ti 3.0 \
    --bolus 1.8 \
    --pvcmethod MLTS

Structural prerequisite

The perfusion pipeline requires a completed Structural Pipeline run for the same subject. It uses the T1 brain, white matter mask, and T1→MNI registration produced by BRC_structural_pipeline.

Arguments¶

Compulsory¶

Argument	Description
`--input_pwi`	Full path to the difference (perfusion-weighted) image
`--input_m0`	Full path to the calibration (M0) image
`--path`	Output directory (absolute path)
`--subject`	Subject identifier

Optional¶

Argument	Default	Description
`--tr`	`3.2`	Repetition time of calibration data (seconds)
`--ti`	`3.0`	Inversion time (seconds)
`--bolus`	`1`	Labelling (bolus) duration (seconds)
`--cgain`	`1`	Relative gain between calibration and ASL image
`--pvcmethod`	`NONE`	Partial volume correction method: `MLTS` or `NONE`
`--name`	`aslMRI`	Output folder name

Processing Steps¶

1. CBF Quantification (`Image_Quantification.sh`)¶

Applies the standard single-compartment kinetic model to the difference image, using TR, TI, bolus duration, and calibration gain to compute absolute CBF in ml/100g/min. The M0 image is used to scale the signal to physiological units.

2. ASL-to-T1 Registration (`ASL_2_T1_Registration.sh`)¶

Registers the mean ASL image to the T1 brain using boundary-based registration (FSL epi_reg style), leveraging the white matter partial volume estimate and WM mask from the structural pipeline for optimal cortical boundary alignment. The registration uses a configurable number of degrees of freedom (DOF) and super-sampling level.

3. Apply Registration (`Apply_Registration.sh`)¶

Propagates the CBF map and ASL data to T1 native space and MNI152 standard space by applying the ASL→T1 affine and the pre-computed T1→MNI warp field in a single applywarp call.

4. Partial Volume Correction (`Partial_Volume_Correction.sh`)¶

When --pvcmethod MLTS is specified, applies the multi-tissue linear series (MLTS) method (mlts_partial_volume_correction.py) using grey matter (GM) and white matter (WM) partial volume estimates from FAST. This corrects for the overestimation of CBF in voxels with mixed tissue types, improving spatial specificity.

5. Data Organisation (`Data_Organization.sh`)¶

Moves processed files into the canonical BRC output hierarchy under <subject>/analysis/aslMRI/.

Kinetic Model Parameters¶

The quantification model assumes single-TI pCASL and applies:

Parameter	Symbol	Default
Repetition time	TR	3.2 s
Inversion time	TI	3.0 s
Bolus duration	τ	1.0 s
Calibration gain	M	1

Modify these via the corresponding command-line arguments to match your acquisition protocol.

Dependencies¶

Tool	Used For
FSL (`flirt`, `applywarp`, `fslmaths`)	Registration, warp application
Python 3	MLTS partial volume correction
Structural pipeline outputs	T1 brain, WM mask, T1→MNI registration
`BRC_GLOBAL_SCR`	Logging
`BRC_PMRI_SCR`	All pipeline sub-scripts

Key Output Files¶

All paths are relative to <path>/<subject>/analysis/aslMRI/.

File	Description
`processed/CBF.nii.gz`	Quantified CBF map in ASL native space
`processed/CBF_2_T1.nii.gz`	CBF map registered to T1 native space
`processed/CBF_2_std.nii.gz`	CBF map in MNI152 standard space
`processed/CBF_PVC.nii.gz`	Partial volume corrected CBF (if MLTS)
`reg/asl2str.mat`	ASL → T1 affine transform
`reg/asl2std_warp.nii.gz`	ASL → MNI warp field
`log/log.txt`	Full processing log

Script Reference¶

Script	Purpose
`ASL_preproc.sh`	Entry point — argument parsing and orchestration
`aslMRI_preproc_part_1.sh`	Processing dispatcher
`Image_Quantification.sh`	Kinetic model CBF quantification
`ASL_2_T1_Registration.sh`	ASL-to-T1 boundary-based registration
`Apply_Registration.sh`	Apply affine + warp to propagate CBF maps
`Partial_Volume_Correction.sh`	PVC wrapper (MLTS method)
`mlts_partial_volume_correction.py`	MLTS partial volume correction implementation
`Data_Organization.sh`	Move outputs into BRC folder hierarchy