Interrater and intrarater reliability of aqueduct CSF flow parameters measured by phase-contrast MRI in healthy adults.

Abstract

Background: Cerebrospinal fluid (CSF) hydrodynamic parameters at the aqueduct of Sylvius are critical for understanding and diagnosing neurodegenerative diseases. This study aims to assess both interrater and intrarater reliability of aqueduct CSF flow parameters measured by phase-contrast magnetic resonance imaging (PC-MRI) across different operators in healthy adults.

Methods: Thirty-eight healthy adult participants were included in the study. Three operators independently quantified aqueduct CSF flow parameters using PC-MRI. Parameters measured included surface (cross-sectional area of aqueduct), net flow per stroke, flow amplitude, peak mean velocity, and stroke volume. Intraclass correlation coefficients [ICC (3,1) for interrater and ICC (1,1) for intrarater reliability] were used to evaluate measurement consistency. Non-parametric tests assessed the impact of velocity encoding (VENC), sex, age, and cardiac period on measurement variability.

Results: Most aqueduct parameters exhibited good to excellent reliability, with ICC (3,1) values ranging from 0.75 to 0.99. Surface demonstrated poor reliability with an ICC (3,1) of 0.37 and ICC (1,1) of 0.33, 0.54 and 0.84. Net flow rate demonstrated moderate consistency with an ICC (3,1) of 0.84 and ICC (1,1) of 0.78, 0.87 and 0.92. VENC significantly increased noise intensity and decreased net flow rate reliability (P<0.01). Gender analysis showed that males had significantly higher surface area (P<0.01), stroke volume (P=0.04), and net flow rate (P<0.01) compared to females.

Conclusions: The study confirmed that most aqueduct CSF flow parameters measured by PC-MRI are reliable both across different operators and in repeated measurements. However, net flow rate and surface area require careful consideration due to lower reliability. These findings provide a valuable reference for the clinical application of CSF hydrodynamic measurements in diagnosing and monitoring neurodegenerative diseases.