Reproducibility of apparent diffusion coefficient measurement in normal prostate peripheral zone at 1.5T MRI

Diffusion-weighted imaging (DWI) is increasingly used to image pathologic processes such as cancer, inflammation, fibrosis or infarction in a wide range of organs [1], [2], [3]. In the prostate, DWI is the corner stone of multiparametric magnetic resonance imaging (MRI) [4, 5]. It is the official ‘dominant sequence’ for detecting prostate cancer in the peripheral zone (PZ) and its role in the transition zone (TZ) has been increased in the last version of the Prostate Imaging-Reporting and Data System (PI-RADS) [6, 7]. Currently, PI-RADS scoring relies on visual inspection of apparent diffusion coefficient (ADC) maps and native trace images obtained with high b-values. Nonetheless, it remains limited by substantial inter-reader variability. Therefore, several groups have suggested using quantitative ADC thresholds to improve the accuracy and reproducibility of visual inspection [8], [9], [10], [11], [12], [13], even if ADC does not directly assess glandular differentiation [14], [15], [16].

However, using quantitative thresholds implies that ADC measurement is highly repeatable and reproducible [17]. Unfortunately, lack of repeatability and reproducibility is a well-known limitation of DWI. ADC values can be influenced not only by hardware-related factors such as gradient systems, coil systems or pulse sequence design, but also by differences in imaging protocols or even by the size and types of the regions-of-interest (ROIs) used for measurement [18], [19], [20], [21], [22], [23], [24], [25], [26], [27]. Despite an abundance of literature on the subject, a recent review performed by the Quantitative Imaging Biomarkers Alliance concluded that available evidence on the repeatability/reproductibility of quantitative biomarkers derived from diffusion-weighted or dynamic contrast-enhanced imaging was scant. It also underlined the need for estimations of the minimal difference that could be distinguished from measurement error, using metrics such as the coefficient of variation (CoV) or the repeatability/reproducibility coefficient (RC) [28]. Another review highlighted the need for assessing ADC reproducibility in abdominopelvic organs using volunteer populations, since variations are far more substantial in vivo than in phantoms [1].

The purpose of this study was to identify the factors of variability in ADC measurement and quantify their influence across different MRI scanners and protocols, in order to better define the minimal ADC differences that can be measured with confidence in the prostate.

留言 (0)

沒有登入
gif