Meeting Banner
Abstract #4153

Reproducibility, Intra- And Inter-Observer Variability of ADC Measurement by Volumetric Segmentation of Bone Marrow in Whole Body Diffusion-Weighted Imaging (WB-DWI)

Sharon L. Giles1, Catherine J. Simpkin1, David John Collins1, 2, Christina Messiou1, Nandita M. deSouza, 12

1MRI Department, Royal Marsden Hospital, Sutton, Surrey, United Kingdom; 2Clinical Magnetic Resonance, Institute of Cancer Research, Sutton, Surrey, United Kingdom

WB-DWI is emerging as a quantitative tool to assess treatment response in myeloma but changes in ADC indicative of response may be relatively small and analysis methods remain without consensus. This prospective study investigated the reproducibility, intra- and inter-observer variability of a volumetric segmentation technique in a cohort of healthy volunteers. Reproducibility of the ADC measurement was between 0.3% and 9.2% and the segmentation technique used provided a consistent and reliable method to generate ADC histograms. It did not introduce additional variation across or between observers and is therefore of potential value in clinical assessment of treatment response in myeloma.

Keywords

abdomen acknowledged acknowledgments acquisition additional allow although apparent assess assessed assessment association audience bandwidth basis biological biomedical bland blocks blood body bone breathing calculating cancer class clinical clinicians coefficient coefficients cohort coil confidence conjunction consensus consistent consistently correlation coverage covered criteria crucial degree dependent diffusion double england error evaluate evaluated extremely free functional funding gall generate gradients grant graphical health healthy highly histograms illustrated illustration inconsistencies indicative institute institution inter interval intra introduce kingdom larger lumbar made mainly marrow matrix narrow observer observers occasion patient pelvis percentage period phantom physicists pixels plot plots position previously process provision quantitative radiographers radiologists recommend registered relatively remaining remains reproducibility reproducible respectively response role royal scanned segmentation segmentations segmented selected separate shape shimming shot skeleton slice slices software spin spine stations statistic studies suppression surface surrey system table target third tool treatment undertaken unknown variability variation variations visible volume volumetric volunteer volunteers whether whilst whole