Biomarkers.

Area of Science:

• Neuroimaging
• Medical Physics
• Radiology

Background:

• Arterial Spin Labelling (ASL) is a non-invasive MRI technique for assessing cerebral blood flow (CBF).
• Barriers to ASL data analysis in resource-constrained settings (RCS) include lack of expertise and high computational costs.
• ASL-MRICloud offers a cloud-based, open-source solution for ASL processing, suitable for RCS.

Purpose of the Study:

• To implement ASL-MRICloud in Google Colab for scalable and cost-effective ASL data analysis.
• To enhance the reproducibility of population studies in RCS.
• To train African researchers in neuroimage analysis through the CONNExIN program.

Main Methods:

• ASL-MRICloud was implemented in Google Colab for automated ASL processing and CBF quantification.
• The Open Science Initiative for Perfusion Imaging ASL (OSIPI-ASL) MRI Challenge dataset was used for initial testing and validation.
• The pipeline was further tested by four CONNExIN teams analyzing data from the PREVENT-AD study.

Main Results:

• Preliminary results from the OSIPI-ASL dataset are available (Figure 1).
• The pipeline, documentation, and team analysis results will be publicly shared on Protocol.io.

Conclusions:

• The study assesses the feasibility of a resource-efficient tool for reproducible ASL data analysis using simulated and real-world data.
• The approach will be shared for wider adoption in RCS to promote inclusive and reproducible imaging research.