Monitoring Carotid Artery Stiffness and Pulse Wave Velocity Using Non-Contrast-Enhanced 4D MR Angiography

Presenting Author(s)

• IM

  Isabel Montón Quesada, MSc

  PhD Student
  University Hospital (CHUV) and University of Lausanne (UNIL)
  Lausanne, Vaud, Switzerland

Primary Author(s)

• IM

  Isabel Montón Quesada, MSc

  PhD Student
  University Hospital (CHUV) and University of Lausanne (UNIL)
  Lausanne, Vaud, Switzerland

Co-Author(s)

• TB

  Thomas Baumgartner, MD

  Resident
  Neurology Service, Cerebrovascular Center, University Hospital of Lausanne (CHUV), Switzerland
• 

  Augustin C. Ogier, PhD

  Post-doctoral fellow
  Lausanne University Hospital (CHUV) and University of Lausanne (UNIL)
  Lausanne, Vaud, Switzerland
• RF

  Robin Ferincz, MSc

  PhD student
  University Hospital (CHUV) and University of Lausanne (UNIL), Switzerland
• JL

  Jean-Baptiste Ledoux, MSc

  Senior Technologist
  Lausanne University Hospital (CHUV)
  Lausanne, Vaud, Switzerland
• 

  Christopher W. Roy, PhD

  Lecturer
  Lausanne University Hospital (CHUV) and University of Lausanne (UNIL)
  Lausanne, Vaud, Switzerland
• JY

  Jerome Yerly, PhD

  Research scientist
  University Hospital (CHUV) and University of Lausanne (UNIL), Vaud, Switzerland
• MS

  Matthias Stuber, PhD

  Professor/Director
  CIBM/CHUV/UNIL
  Lausanne, Switzerland
• LH

  Lorenz Hirt, MD

  Associate Professor
  Neurology Service, Cerebrovascular Center, University Hospital of Lausanne (CHUV), Switzerland
• 

  Ruud B. van Heeswijk, PhD, FSCMR

  Senior lecturer
  Lausanne University Hospital (CHUV) and University of Lausanne (UNIL)
  Lausanne, Vaud, Switzerland

Background: Static common carotid artery (CCA) anatomy and atherosclerotic plaque can be studied with magnetic resonance imaging (MRI) [1], but stiffness and other dynamic CCA properties are routinely evaluated using ultrasound (US) by tracking the diameter changes in CCA [2]. We therefore aimed to develop an operator-independent 3D MRI technique to measure the CCA diameter change throughout the cardiac cycle and to provide stiffness-related parameters that could be complementary to the existing US toolset. To this end, we implemented free-running (without any triggering or gating) CCA MR angiography (MRA) without contrast agent injection and characterized the technique in junior and senior participants.

Methods: N=18 healthy volunteers were included and divided into two age subgroups: a junior cohort (n=9, 3F, 29±5Y) and a senior cohort (n=9, 6F, 61±7Y). Their CCA was scanned using both a US system (EPIQ Elite, Philips) (Fig. 1A) and a 3T MR scanner (PrismaFit, Siemens) using a free-running 3D radial GRE acquisition [3, 4] and slab-selective water excitation (flip angle 8º, pixel size (0.6mm)³, duration 10min). Both scans were performed twice to assess repeatability. MRI data were retrospectively sorted into 80 ms cardiac phases using pulse oximetry and reconstructed using compressed sensing [5] (Fig. 1B, 2).

CCA stiffness was evaluated as β=ln(P_s/P_d)/((D_s−D_d)/D_d), where P<sub>s and P<sub>d are the systolic and diastolic pressures (measured with an arm cuff), and D_s and D_d are the systolic and diastolic lumen diameters, measured manually for US (Fig. 1C,E) and semi-automatically for MRA (Fig. 1D,F). Local pulse wave velocity [6] was computed as PWV=√(β×P_s/2ρ), where ρ is the blood density.

Paired Student’s t-tests were used to compare β and PWV between age groups and between imaging techniques. Repeatability was quantified as the intraclass correlation coefficient (ICC) and coefficient of variation (CV) on both the diastolic and systolic diameters.

Results: As expected, the CCA β measured with MRA (β_{junior_MRA}=3.8±0.9 vs β_{senior_MRA}=5.4±1.0, p=0.002) and US (β_{junior_ US}=3.5±0.7 vs β_{senior_ US}=7.2±3.0, p=0.005) was significantly higher in the senior cohort (Fig. 2A). When comparing modalities, there was no significant US-MRA difference in the junior (p=0.5) or senior (p=0.08) cohorts. The PWV was significantly higher for the senior cohort with both MRA (PWV_{junior_MRA}=5.2±0.7m/s vs PWV_{senior_MRA}=6.5±0.9m/s, p=0.0005) and US (PWV_{junior_US}=5.0±0.5m/s vs PWV_{senior_US}=7.4±1.7 m/s, p=0.003, Fig. 2B). No significant PWV differences were observed between the two modalities. MRA presented excellent repeatability with higher ICC than US (ICC_MRA=0.94 vs. ICC_US=0.67) and lower CV (CV_MRA=1.7% vs. CV_US=3.8%).

Conclusion: Free-running 4D MRA enables the repeatable non-invasive assessment of CCA anatomy and dynamics, offering precision comparable to US and enabling the detection of age-related increases in arterial β and PWV, demonstrating its potential as a complementary tool for 3D vascular aging assessment.

Figure 1. Overview of US and MRI acquisition and analysis pipelines. (A) Acquisition of M-mode US images. (B) Continuous acquisition of MRA data using a segmented 3D golden-angle top-to-bottom phyllotaxis trajectory and k-space cardiac binning into N bins of 80 ms using the pulse oximeter signal. (C) Manual delineation of the interface between the vessel wall and the blood pool in the case of US. (D) Semiautomatic detection of the vessel wall using a full-width half-maximum (FWHM) approach. (E) For US, computation of the systolic and diastolic diameters by subtracting the previously delineated boundaries. (F) Detection of the systolic and diastolic diameters in the case of MRA.
Figure 2. Three example orthogonal anatomical views of the neck and carotid arteries in the MRA in a junior participant together with a 3D rendering of the carotid arteries. While the arterial blood has a bright signal, the lipid signals are largely suppressed by the water excitation.
Figure 3. Stiffness-related measurements for both modalities and cohorts in the common carotid arteries. (A) Stiffness index (β) measurements showed no statistical significance between modalities while there was an age-related increase of β when comparing the junior and senior cohorts (p_MRA = 0.002 and p_US = 0.005).(B) Local pulse wave velocity (PWV) values were significantly higher for the senior cohort in both modalities (p_MRA = 0.0005 and p_US = 0.003) while there were no statistical differences between MRA and US.