Intra-tumoral susceptibility signal: A post-processing technique for objective grading of astrocytoma with susceptibility-weighted imaging

Background: Susceptibility-weighted imaging (SWI) is sensitive to the accumulation of paramagnetic substances, such as hemorrhage and increased venous vasculature, both being frequently found in high-grade tumors. The purpose of this retrospective study is to differentiate high-grade and low-grade astrocytoma by objectively measuring quantitative intra-tumoral susceptibility signals (qITSS) on SWI. Methods: Precontrast SWI and 3D contrast-enhanced T1WI of 65 patients with astrocytoma were collected at 1.5 Tesla. All tumors were histologically confirmed and classified into two groups: high grade (WHO grade III and IV, n=50) and low grade (WHO grade II, n=15). After manual delineation of the tumor on T1WI, normalized contrast (NC) was calculated voxel by voxel within the tumor by using the concept of contrast to noise ratio. Thresholding on NC was applied to detect qITSS, and the volumetric percentage of qITSS can be obtained for each tumor. Two-sample t-test was applied to examine significant difference of qITSS percentage between high-grade and low-grade astrocytoma for different NC thresholds, ranging from 4 to 20. Receiver operating characteristic analysis was performed to evaluate the performance of differentiation. Results: P value was less than 0.01 for a large range of NC thresholds [4–20], reflecting significant difference of qITSS percentage between high-grade and low-grade astrocytoma. The area under the receiver operating characteristic curve was larger than 0.9 at NC thresholds from 8 to 16 and peaks at 0.949 with a NC threshold of 14. It was shown that astrocytoma grading by qITSS percentage is successful for a wide range of NC threshold, demonstrating robustness on threshold selection. Conclusions: Without relying on the selection of slice position and at the same time providing objective identification of hypointense signal in SWI, the qITSS percentage can be used to distinguish high-grade and low-grade astrocytoma reliably. © Quantitative Imaging in Medicine and Surgery. All rights reserved.