<scp>MR</scp> Passive Elastography: From Seismic Noise Tomography to Brain Tumors Characterization

ABSTRACT Background Noise correlation tomography has revolutionized seismology by enabling constant monitoring of our planet. Considering the brain as a system full of noise, the same algorithm might be adapted to monitor the brain using MRI. Purpose To apply the MR passive elastography sequence to patients diagnosed with brain tumors. Study Type Prospective. Population Four healthy volunteers and 18 patients (11 women and 7 men) with histopathology‐confirmed brain tumors. There were 12 malignant tumors (10 high‐grade gliomas, 2 metastases) and 6 had benign tumors (2 low‐grade gliomas, 4 meningiomas). Field Strength/Sequence A modified T2‐weighted spin‐echo sequence with motion encoded in one spatial direction at 3.0 T. Assessment Shear wavelength estimated from the passive elastography sequence was compared to the shear modulus measured with intraoperative ultrasound. Statistical Tests Agreement of tumor mechanical properties measurements was assessed using Pearson correlation and Bland–Altman plots. The ability to distinguish malignant from benign tumors was evaluated using receiver operating characteristic curve analysis. Significance was considered at a p value of 0.05. Results Tumor wavelength values obtained with passive elastography were significantly correlated with ultrasound elastography ( r = 0.59). Moreover, by calculating the ratio between normal brain and tumor, passive elastography was able to distinguish malignant tumors (shear wavelength of 0.26 m) from benign tumors (shear wavelength of 0.29 m) (AUC = 0.93). Data Conclusions Intracranial passive elastography is a feasible technique that does not require an external vibrating driver and has the potential to provide valuable information about tumors. Evidence Level 2. Technical Efficacy Stage 2.