Arterial Spin Labeling Perfusion MRI Signal Processing Through Traditional Methods and Machine Learning.

Area of Science:

• Radiology and Imaging
• Neuroscience
• Medical Physics

Background:

• Arterial spin labeling (ASL) perfusion MRI is a non-invasive method for cerebral blood flow (CBF) quantification.
• ASL MRI signal is influenced by factors like arterial blood volume, signal decay, physiological fluctuations, and head motion.
• While some factors are managed by ASL MRI sequences, others require robust post-processing strategies.

Purpose of the Study:

• To provide a comprehensive review of ASL MRI data processing.
• To introduce typical methods used in ASL MRI data processing pipelines.
• To briefly discuss both traditional heuristic and recent machine learning-based approaches.

Main Methods:

• Overview of ASL MRI data processing pipeline steps.
• Introduction to traditional heuristic model-based post-processing methods.
• Brief introduction to machine learning-based approaches for ASL MRI data analysis.

Main Results:

• ASL MRI requires sophisticated post-processing to address signal variations and improve CBF quantification.
• A range of heuristic methods have been developed over decades for ASL MRI data analysis.
• Emerging machine learning techniques offer potential advancements in ASL MRI data processing.

Conclusions:

• Effective post-processing is crucial for accurate ASL MRI-based CBF mapping.
• This review provides a foundational understanding of ASL MRI processing, encompassing traditional and novel methods.
• Further research into advanced post-processing techniques, including machine learning, is warranted for optimizing ASL MRI applications.