Meeting Banner
Abstract #3387

Diffusional Kurtosis Imaging of Prostate Cancer: Effect of B-Values and Noise Compensation on Quantitative Parameters, Relative Contrast, and Short-Term Repeatability

Edward M. Lawrence1, Andrew N. Priest1, Tristan Barrett1, Anne Warren2, Ferdia A. Gallagher1, Vincent J. Gnanapragasam3, Evis Sala1

1Radiology, Addenbrooke's Hospital, Cambridge, Cambridgeshire, United Kingdom; 2Histopathology, Addenbrooke's Hospital, Cambridge, Cambridgeshire, United Kingdom; 3Urology, Addenbrooke's Hospital, Cambridge, Cambridgeshire, United Kingdom

Diffusional kurtosis imaging (DKI) attempts to more accurately quantify water movement by taking into account the possibility of a non-Gaussian distribution. Research into the potential of DKI in the setting of body MRI has been limited and the purpose of this study was to evaluate the effect of b-values and noise compensation on the evaluation of prostate cancer using DKI. Our results showed that DKI is a robust technique that allows for significant differentiation of cancerous regions, as established by whole-mount pathology, regardless of DKI method or tumor location. The highest relative contrast was gained using a high b-value of 1500 s/mm2.

Abstract PDF Presentation Video

Keywords

accounting acquisition additional additionally advanced aggressiveness anatomical apparent approved approximation array assessment asset attempts audience better biopsy body cancer channel clinically clinicians coefficient coil comp compensating compensation confirmed consecutive contrast correction decrease decreased dependence detection deviations differentiation diffusion diffusional distribution drawn dual eighteen equation evaluate evaluated evaluation excluded experience experienced failure feasibility fitting floor gained getting good highest hospital identical impact improvement included index influence initial interestingly investigate investigated investigation kingdom kurtosis landmarks location locations made maps matched matrix measure measured model mount movement necessary negative noise nonlinear note outlined paired pathology patient patients peripheral physicists preliminary priest prospective prostate pulses quantification quantify quantitative quantitatively radiology reader recruited regardless remained repeatability respectively retest robust scanner scheduled separately setting short significantly slice spin subgroup table target term tissue trace transition tumor tumors type urology variation warren water whole year zero zone