Whole-tumor histogram analysis of apparent diffusion coefficient in differentiating intracranial solitary fibrous tumor/hemangiopericytoma from angiomatous meningioma

Publication date: Available online 22 January 2019

Source: European Journal of Radiology

Author(s): Wenle He, Xiang Xiao, Xiaodan Li, Yihao Guo, Liuji Guo, Xiaomin Liu, Yikai Xu, Jun Zhou, Yuankui Wu

Abstract

Purpose

To assess the role of histogram analysis of apparent diffusion coefficient (ADC) maps based on whole-tumor in differentiating intracranial solitary fibrous tumor/hemangiopericytoma (SFT/HPC) from angiomatous meningioma (AM).

Materials and methods

Pathologically confirmed intracranial SFT/HPC (n = 15) and AM (n = 20) were retrospectively collected and their clinical and conventional MRI features were analyzed. Diffusion-weighted (DW) images (b = 0 and 1000s/mm²) were processed with the mono-exponential model. Regions of interest covering the whole tumor were drawn on all slices of the ADC maps to obtain histogram parameters, including mean ADC (ADCmean), median ADC (ADCmedian), maximum ADC (ADCmax), minimum ADC (ADCmin), skewness and kurtosis, as well as the 5^th, 10^th, 25^th, 75^th, 90^th and 95^th percentile ADC (ADC5, ADC10, ADC25, ADC75, ADC90 and ADC95). Differences of histogram parameters between SFT/HPC and AM were compared using Mann-Whitney U test. Receiver operating characteristic (ROC) curves was used to determine the diagnostic performance.

Results

The ADCmin (P=0.001) and ADC5 (P = 0.045) were significantly lower in SFT/HPCs than in AMs, while no significant difference was found in sex, age, conventional MRI features or any other histogram parameters between the two entities (P=0.051-1.000). ADCmin showed the best diagnostic performance (area under curves [AUC], 0.86; sensitivity, 81.3%; specificity, 83.3%) in differentiating SFT/HPC from AM with optimal cutoff value being 569.00 × 10^-6  mm²/s, followed by ADC5 (AUC, 0.72; sensitivity, 68.8%; specificity, 75 %) with optimal cutoff value being 781.97 × 10^-6  mm²/s.

Conclusion

SFT/HPC and AM share similar conventional MR appearances. Whole-tumor histogram analysis of ADC maps may be a useful tool for differential diagnosis, with ADCmin and ADC5 being potential parameters.

from #Head and Neck by Sfakianakis via simeraentaxei on Inoreader http://bit.ly/2WbS29j