Quantitative basal CBF and CBF fMRI of rhesus monkeys using three-coil continuous arterial spin labeling

Xiaodong Zhang, Tsukasa Nagaoka, Edward J. Auerbach, Robbie Champion, Lei Zhou, Xiaoping Hu, Timothy Q. DuongNeuroImage 2007 · DOI

TL;DR

This paper introduces a novel three-coil arterial spin labeling (ASL) MRI technique for high-resolution, quantitative cerebral blood flow (CBF) measurements in rhesus monkeys. This advancement enables simultaneous CBF and BOLD fMRI, paving the way for better understanding of brain function and neurological diseases in non-human primates.

What Problem Does This Paper Address?

Accurately measuring cerebral blood flow (CBF) non-invasively is crucial for understanding brain physiology and pathology. While MRI techniques like Blood-Oxygen-Level-Dependent (BOLD) fMRI are widely used, they indirectly reflect neural activity through changes in blood oxygenation. Arterial Spin Labeling (ASL) offers a more direct measure of CBF, but achieving high temporal and spatial resolution, especially in non-human primates, has been challenging. Existing ASL methods often struggle with signal-to-noise ratio (SNR) and resolution, limiting their application for detailed functional studies. This paper aims to overcome these limitations by developing and validating a specialized ASL setup for rhesus monkeys, enabling quantitative CBF and CBF-fMRI with improved resolution and SNR.

Methods

The core innovation is a custom-built three-coil continuous ASL (CASL) system integrated with a Siemens 3T Trio scanner. This setup includes:

  • A separate neck labeling coil: This coil is dedicated to magnetically labeling arterial blood in the neck, improving the efficiency and specificity of the ASL signal. It also helps to eliminate magnetization transfer effects that can contaminate the signal.
  • A receive-only surface coil for brain imaging: This coil is designed for optimal signal reception from the monkey brain, enhancing SNR and spatial resolution.
  • Active decoupling circuits: These circuits ensure that the different coils do not interfere with each other, crucial for accurate signal acquisition and preventing unwanted excitation of brain tissue.

The system was optimized for labeling power (2 W), labeling efficiency (92±2%), and post-labeling delay (0.8 s). Quantitative CBF was measured with 1.5-mm isotropic resolution in 3 minutes. CBF-fMRI and BOLD-fMRI were acquired simultaneously with an 8-second temporal resolution during hypercapnic (5% CO2) and hyperoxic (100% O2) challenges.

Datasets

The study used six rhesus monkeys as subjects. No specific publicly available datasets were used for training or evaluation; rather, the data was acquired directly from these animal subjects.

Key Insights and Ideas

  • The power of dedicated coils: Using a separate neck coil for ASL labeling, combined with a specialized brain receive coil, significantly boosts SNR and allows for higher spatial resolution than single-coil approaches. This is particularly important for small animal brains.
  • Optimizing ASL parameters is critical: The paper meticulously details the process of optimizing labeling power, efficiency, and post-labeling delay, demonstrating how these parameters directly impact the quality and interpretability of CBF measurements.
  • CBF-fMRI complements BOLD: The study shows that while hypercapnia increases both CBF and BOLD signals, hyperoxia leads to decreased CBF and increased BOLD. This highlights the complementary information provided by CBF-fMRI, offering a more direct measure of metabolic demand than BOLD alone.
  • High-resolution CBF is achievable: The ability to acquire quantitative CBF maps with 1.5-mm isotropic resolution in just 3 minutes is a significant step forward for primate neuroimaging.

Key Results

  • Quantitative Basal CBF:
    • Gray Matter (GM) CBF: 104±3 ml/100 g/min
    • White Matter (WM) CBF: 45±6 ml/100 g/min
    • GM/WM Ratio: 2.3±0.2
  • CBF-fMRI Responses:
    • Hypercapnia (5% CO2): +59±10% (GM), +37±4% (WM)
    • Hyperoxia (100% O2): -9.4±2% (GM), -3.9±2.6% (WM)
  • BOLD-fMRI Responses:
    • Hypercapnia (5% CO2): +2.0±0.4% (GM), +1.2±0.4% (WM)
    • Hyperoxia (100% O2): +2.4±0.7% (GM), +0.8±0.2% (WM)
  • Technical Achievements: Achieved 1.5-mm isotropic resolution for quantitative CBF in 3 minutes and 8-second temporal resolution for combined CBF/BOLD fMRI.

Limitations and Future Work

  • Anesthesia Effects: The study was conducted under isoflurane anesthesia, a known vasodilator. The authors acknowledge that this can influence basal CBF and vascular reactivity, making direct comparisons with awake conditions or other anesthetics challenging. Future work should investigate CBF and BOLD responses under different anesthetic regimens and in awake monkeys.
  • Water Exchange and Transit Time: While efforts were made to account for arterial transit time (ATT), the authors note that water exchange effects can still influence absolute CBF values, particularly in white matter. Further research into more sophisticated quantification models is suggested.
  • Spatial and Temporal Resolution Improvements: While significant improvements were made, the authors suggest that further enhancements in spatial and temporal resolution are desirable and expected with advancements in segmented EPI acquisition and parallel imaging techniques.
  • Hardware Implementation: The custom hardware developed for this study could be adapted for other MRI vendors, suggesting a path for broader adoption of this technique.

Summary updated

BibTeX
@article{zhang2007quantitative,
  author = {Xiaodong Zhang and Tsukasa Nagaoka and Edward J. Auerbach and Robbie Champion and Lei Zhou and Xiaoping Hu and Timothy Q. Duong},
  title = {{Quantitative basal CBF and CBF fMRI of rhesus monkeys using three-coil continuous arterial spin labeling}},
  journal = {NeuroImage},
  year = {2007},
  doi = {10.1016/j.neuroimage.2006.10.011},
}
Figure