Diffusion kurtosis imaging study of prostate cancer: preliminary findings

J Magn Reson Imaging. 2014 Sep;40(3):723-9. doi: 10.1002/jmri.24379. Epub 2013 Oct 31.

Abstract

Purpose: To evaluate the differences in parameters of diffusion kurtosis imaging (DKI) between prostate cancer, benign prostatic hyperplasia (BPH), and benign peripheral zone (PZ).

Materials and methods: Twenty-four foci of prostate cancer, 41 BPH nodules (14 stromal and 27 nonstromal hyperplasia), and 20 benign PZ from 20 patients who underwent radical prostatectomy were investigated. Diffusion-weighted imaging (DWI) was performed using 11 b-values (0-1500 s/mm(2) ). DKI model relates DWI signal decay to parameters that reflect non-Gaussian diffusion coefficient (D) and deviations from normal distribution (K). A mixed model analysis of variance and receiver operating characteristic (ROC) analyses were performed to assess the statistical significance of the metrics of DKI and apparent diffusion coefficient (ADC).

Results: K was significantly higher in prostate cancer and stromal BPH than in benign PZ (1.19 ± 0.24 and 0.99 ± 0.28 versus 0.63 ± 0.23, P < 0.001 and P < 0.001, respectively). K showed a trend toward higher levels in prostate cancer than in stromal BPH (1.19 ± 0.24 versus 0.99 ± 0.28, P = 0.051). On the ROC analyses, a significant difference in area under the curve was not observed between K and ADC, however, K showed the highest sensitivity among three parameters.

Conclusion: DKI may contribute to the imaging diagnosis of prostate cancer, especially in the differential diagnosis of prostate cancer and BPH.

Keywords: benign prostatic hyperplasia; diffusion; kurtosis; magnetic resonance imaging; prostate cancer.

MeSH terms

  • Aged
  • Diffusion Magnetic Resonance Imaging / methods
  • Humans
  • Male
  • Neoplasm Grading
  • Neoplasm Staging
  • Prostatectomy
  • Prostatic Hyperplasia / pathology*
  • Prostatic Hyperplasia / surgery
  • Prostatic Neoplasms / pathology*
  • Prostatic Neoplasms / surgery
  • Retrospective Studies