Meeting Banner
Abstract #3398

Optimization of B-Value Distribution for Biexponential DWI of Normal Prostate

Ivan Jambor1, 2, Jukka Jrvinen1, 3, Hannu J. Aronen1, 3, Jani Saunavaara3, Harri Merisaari4, Tommi Kauko5, Ronald Borra1, 3, Marko Pesola, 13

1Department of Diagnostic Radiology, University of Turku, Turku, Finland; 22nd Department of Radiology, Comenius University and St. Elisabeth Oncology Institute, Bratislava, Slovakia; 3Medical Imaging Centre of Southwest Finland, Turku University Hospital, Turku, Finland; 4Turku PET centre, University of Turku, Turku, Finland; 5Department of Biostatistics, University of Turku, Turku, Finland

The optimal b-value distribution for biexponential DWI of normal prostate was determined using using Monte-Carlo simulations and in-vivo measurements. Eight healthy volunteers underwent four repeated 3T prostate DWI scans using both 16 equally distributed b-values and an optimized b-value distribution obtained from the simulations. The b-value distributions were compared in terms of measurement reliability and repeatability using Shrout-Fleiss analysis. The optimal b-value distribution was found to be a clustered distribution with b-values concentrated in the low, mid and high ranges and was shown to improve the estimation quality of biexponential DWI parameters of in-vivo experiments.

Abstract PDF Presentation Video

Keywords

according accuracy added addition aimed apparent better biostatistics calculation cancer characterization class clustered clusters coefficient coefficients components concentrated consecutively consequently correlation criterion curve curves days decay decreased described detection determine diagnostic diffusion distributed distribution distributions eight entire equal equally error estimation evaluate evenly examinations exponential fast fewer findings fitted formed fraction function generate generated gland healthy hospital improve improves in vivo individual initial institute intensity intra intraclass iteration iterative literature matrix medical mono moved noise oncology optimal optimization particularly process produced prostate quality radiology randomly reliability repeatability repeated represents selected separate seven shot simulated simulations slice slow southwest starting step stepwise terms tissue underwent valuable volunteers week years