Temporal clustering, tissue composition, and total variation for mapping oxygen extraction fraction using QSM and...

Area of Science:

• Neuroimaging
• Medical Physics
• Quantitative Susceptibility Mapping

Background:

• Quantitative susceptibility mapping plus quantitative blood oxygen level-dependent magnitude (QSM+qBOLD or QQ) is used for mapping oxygen extraction fraction (OEF) and cerebral metabolic rate of oxygen (CMRO2).
• Previous methods like cluster analysis of time evolution (CAT) have limitations in accuracy and robustness.

Purpose of the Study:

• To improve the accuracy and robustness of QQ-based OEF and CMRO2 mapping.
• To introduce a novel method, CCTV mapping, incorporating tissue composition and total variation.

Main Methods:

• Developed CCTV mapping by integrating tissue-type information into temporal clustering and total variation.
• Acquired 3D multi-echo gradient echo and arterial spin labeling images from healthy subjects and ischemic stroke patients.
• Reconstructed QQ-based OEF and CMRO2 using CAT, CAT+TV (CATV), and CCTV, comparing results via statistical analysis.

Main Results:

• CCTV demonstrated superior accuracy and precision in OEF mapping compared to CAT and CATV in simulations.
• In healthy subjects, CCTV resulted in less noisy and more uniform QQ-based OEF.
• In stroke patients, CCTV yielded a significantly higher contrast-to-noise ratio for OEF between lesions and healthy tissue.

Conclusions:

• CCTV mapping significantly enhances the robustness of QQ-based OEF measurements against image noise.
• The proposed method shows promise for more reliable brain oxygenation assessment in clinical settings, especially for stroke patients.