Interstudy reproducibility of three‐dimensional volume‐selective fast spin echo magnetic resonance for quantifying carotid artery wall volume

Anitha Varghese, Lindsey A. Crowe, Raad Mohiaddin, Peter Gatehouse, Guang Yang, David Nott, James McCall, David Firmin, Dudley J. PennellJournal of Magnetic Resonance Imaging 2005 · DOI

TL;DR

This paper introduces and validates a new 3D MRI technique for precisely measuring carotid artery wall volume, a key indicator of atherosclerosis. The study demonstrates excellent inter-scan reproducibility, suggesting this method is robust enough for tracking disease progression over time with relatively small patient cohorts.

What Problem Does This Paper Address?

Atherosclerotic disease, particularly affecting the carotid arteries, is a major contributor to mortality and morbidity, often leading to stroke. Early detection and monitoring of this disease are crucial for effective prevention and treatment. While conventional angiography focuses on lumen narrowing, atherosclerosis also involves arterial wall thickening and remodeling before significant lumen reduction occurs. Therefore, accurately measuring the volume of the arterial wall itself is a more sensitive marker for early disease and progression. Existing MRI techniques for this purpose, while promising, have faced challenges with reproducibility, limiting their utility for longitudinal studies that track subtle changes over time. This paper addresses the need for a reproducible MRI method to quantify carotid artery wall volume, enabling more accurate tracking of atherosclerosis progression.

Methods

The researchers developed and evaluated a three-dimensional (3D) volume-selective fast spin echo (FSE) magnetic resonance (MR) technique. This method is an advancement over previous 2D FSE techniques, offering greater coverage of the arterial volume in a similar scan time. The core of the measurement involves calculating the total carotid arterial wall volume (TWV) by subtracting the total carotid lumen volume from the total outer carotid vessel volume. This is achieved by acquiring cross-sectional MR images of the carotid artery.

Key aspects of their methodology include:

  • 3D Volume-Selective FSE: This sequence allows for efficient acquisition of a 3D dataset covering a significant length of the carotid artery (up to 56 mm in theory, practically around 28 mm).
  • Customized Coil and Cushion: A specialized carotid coil and a custom-made head and neck cushion were used to ensure precise positioning and patient comfort, minimizing motion artifacts.
  • Image Analysis: TWV was calculated by a single observer using dedicated software (CMRTools). Slices were matched between scans based on their distance from the carotid bifurcation, and only those with adequate signal-to-noise ratio (SNR) were included. T2-weighted images were used to help differentiate vessel wall from flow artifacts.
  • Reproducibility Assessment: Interstudy reproducibility was assessed by having 10 subjects undergo two scans within a short timeframe (1 hour to 4 days). Statistical analyses included paired t-tests, calculation of the coefficient of variation (COV), intraclass correlation coefficient (r), and Bland-Altman plots.

Datasets

The study utilized data from 10 human subjects with diagnosed carotid artery atherosclerosis, confirmed by prior Doppler ultrasonography. These subjects were scanned twice. No new datasets were introduced or made publicly available.

Key Insights and Ideas

  • Early Atherosclerosis Detection: The ability to measure arterial wall volume, rather than just lumen narrowing, allows for the detection of atherosclerosis at an earlier stage, reflecting the phenomenon of arterial remodeling.
  • 3D Advantage: The 3D volume-selective FSE technique offers a significant advantage in terms of coverage and efficiency compared to previous 2D methods.
  • Robust Reproducibility is Key: For longitudinal studies tracking disease progression, consistent and reliable measurements are paramount. This study's focus on interstudy reproducibility directly addresses this need.
  • Carotid-Coronary Link: The paper highlights the strong pathological links between coronary artery disease and stroke, emphasizing the importance of studying carotid atherosclerosis as a proxy for broader cardiovascular health.

Key Results

  • High Reproducibility: The mean carotid TWV was very similar between the two scans (741 mm³ and 734 mm³), with no statistically significant difference (P = 0.5).
  • Low Variability: The standard deviation of the differences between measurements was 33 mm³, resulting in a coefficient of variation (COV) of 4.4%. This indicates a high degree of consistency.
  • Strong Correlation: The correlation between the two scans was excellent, with an intraclass correlation coefficient (r) of 0.99.
  • Narrow Limits of Agreement: Bland-Altman plots showed narrow limits of agreement, further supporting the technique's reproducibility.
  • Sample Size Efficiency: Power calculations indicated that a sample size of 16 patients would be sufficient to detect a significant difference in plaque volume over time (with 80% power at P < 0.05) based on the observed variability.
  • Scan Time: Each CMR study took approximately 20 minutes.

Limitations and Future Work

  • Single Observer: The reproducibility was assessed by a single observer. Future work could involve assessing inter-observer variability.
  • Limited Coverage: While the technique offers 3D coverage, the practical imaging range along the carotid artery was limited (average 28 mm). This variability in coverage might restrict its use to serial evaluations rather than isolated assessments.
  • Future Modifications: The authors suggest several future improvements:
  • Developing a velocity-sensitive phase reconstruction method to identify and remove residual signal from slow-flowing blood, potentially improving accuracy.
  • Incorporating navigator elements (e.g., at the base of the tongue) to further reduce motion artifacts from swallowing and head movement.
  • Developing automated software for interpolating missing slices to increase coverage.
  • Exploring the use of multispectral CMR (using different contrast weightings) and contrast agents (like gadolinium and USPIOs) to identify more detailed plaque characteristics such as fibrous cap rupture, neovasculature, and macrophage density.

Summary updated

Abstract

Volumetric analysis with CMR of carotid artery plaques using a three-dimensional volume-selective FSE is efficient with good interstudy reproducibility, and is well suited for longitudinal studies of progression of carotid atheroma with reasonable sample sizes.
BibTeX
@article{varghese2005interstudy,
  author = {Anitha Varghese and Lindsey A. Crowe and Raad Mohiaddin and Peter Gatehouse and Guang Yang and David Nott and James McCall and David Firmin and Dudley J. Pennell},
  title = {{Interstudy reproducibility of three‐dimensional volume‐selective fast spin echo magnetic resonance for quantifying carotid artery wall volume}},
  journal = {Journal of Magnetic Resonance Imaging},
  year = {2005},
  doi = {10.1002/jmri.20249},
}
Figure