Blood Flow Velocity Analysis in Cerebral Perforating Arteries on 7T 2D Phase Contrast MRI with an Open-Source...

Area of Science:

• Neuroimaging
• Biomedical Engineering
• Medical Physics

Background:

• Cerebral perforating arteries' stiffness, assessed by velocity pulsatility index (PI) from 2D PC-MRI, is linked to cerebrovascular diseases.
• Current analysis tools for 2D PC-MRI in small cerebral vessels are lacking, hindering clinical research and application.

Purpose of the Study:

• To introduce Small vessEL MArker (SELMA), a novel, user-friendly, open-source software for analyzing 2D PC-MRI data in cerebral perforating arteries.
• To validate SELMA's performance and assess its reliability for quantifying blood flow velocity and PI in small cerebral vessels.

Main Methods:

• SELMA's analysis algorithm was validated against existing data using Bland-Altman analysis.
• Inter-rater reliability was assessed using 2D PC-MRI data from 60 participants across 8 sites and 3 MRI vendors.
• Key metrics analyzed included mean velocity (vmean), velocity PI, and number of detected vessels (Ndetected).

Main Results:

• SELMA demonstrated high agreement with previous results for vmean (mean difference ± SD: 0.1 ± 0.8 cm/s) and velocity PI (0.01 ± 0.1).
• High inter-rater reliability was observed, with Dice Similarity Coefficient of 0.91 for regions of interest and intra-class coefficients of 0.92 (Ndetected), 0.84 (vmean), and 0.85 (velocity PI).
• Differences in outcome measures were greater between sites than vendors, highlighting challenges in harmonizing 2D PC-MRI sequences.

Conclusions:

• SELMA is a consistent, user-friendly, and reliable open-source tool for analyzing 2D PC-MRI data in small cerebral vessels.
• The software facilitates objective assessment of blood flow dynamics, potentially improving the understanding and diagnosis of cerebrovascular diseases.
• Further efforts are needed to harmonize 2D PC-MRI acquisition protocols across different sites to improve data consistency.