Abstract
Objectives: To evaluate the diagnostic value of intravoxel incoherent motion (IVIM) and diffusional kurtosis imaging (DKI) parameters in nasal or sinonasal squamous cell carcinoma (SCC) patients to determine local control/failure. Methods: Twenty-eight patients were evaluated. MR acquisition used single-shot spin-echo EPI with 12 b-values. Quantitative parameters (mean value, 25th, 50th and 75th percentiles) of IVIM (perfusion fraction f, pseudo-diffusion coefficient D*, and true-diffusion coefficient D), DKI (kurtosis value K, kurtosis corrected diffusion coefficient Dk) and apparent diffusion coefficient (ADC) were calculated. Parameter values at both the pretreatment and early-treatment period, and the percentage change between these two periods were obtained. Results: Multivariate logistic regression analysis: the percentage changes of D (mean, 25th, 50th, 75th), K (mean, 50th, 75th), Dk (mean, 25th, 50th), and ADC (mean, 25th, 50th) were predictors of local control. ROC curve analysis: the parameter with the highest accuracy = the percentage change of D value with the histogram 25th percentile (0.93 diagnostic accuracy). Multivariate Cox regression analyses: the percentage changes of D (mean, 25th, 50th), K (mean, 50th, 75th), Dk (mean, 25th, 50th) and ADC (mean, 25th, 50th) are predictors. Conclusions: IVIM and DKI parameters, especially the D-value’s histogram 25th percentile, are useful for predicting local control. Key Points: • Noninvasive assessment of treatment outcome in SCC patients was achieved using IVIM/DKI. • Several IVIM and DKI parameters can predict the local control. • Especially, the D-value’s histogram 25th percentile has high diagnostic accuracy.
Author supplied keywords
Cite
CITATION STYLE
Fujima, N., Yoshida, D., Sakashita, T., Homma, A., Tsukahara, A., Shimizu, Y., … Shirato, H. (2017). Prediction of the treatment outcome using intravoxel incoherent motion and diffusional kurtosis imaging in nasal or sinonasal squamous cell carcinoma patients. European Radiology, 27(3), 956–965. https://doi.org/10.1007/s00330-016-4440-1
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.