Related Experiment Video
Updated: May 13, 2026
Modeling Posthemorrhagic Hydrocephalus of Prematurity in Rats
Published on: March 28, 2025
Pilot Evaluation of Rapid 2D Brain MR Elastography in Idiopathic Normal Pressure Hydrocephalus
Pranjal Rai1, Matthew C Murphy2, Petrice M Cogswell2
1From the Department of Radiology (P.R., M.C.M., P.M.C., Y.L., C.R.J., R.L.E., J.H.), Information Technology (M.L.S.), Neurology (J.G.-R.) and Neurologic Surgery (B.D.E.), Mayo Clinic, 200 1st Street Southwest, Rochester, MN 55902. raipranjal2@gmail.com.
Background:
Magnetic Resonance Elastography (MRE) of the brain provides quantitative biomarkers of mechanical properties of brain tissue. Prior studies have shown that an advanced 3D MRE can differentiate idiopathic normal pressure hydrocephalus (iNPH), from other neurodegenerative disorders. In this pilot study, we assessed whether a simplified 2D MRE protocol, that is more widely available, could detect stiffness differences between iNPH and non-iNPH cognitively impaired patients, thus providing preliminary diagnostic discrimination, while using a previously developed 3D MRE classifier as a secondary exploratory comparator.
Methods:
In this prospective study, 10 patients with clinically diagnosed iNPH (iNPH-group) and 10 cognitively impaired patients due to non-iNPH causes (CI-group), underwent 2D brain MRE between September 2024 and April 2025. Median stiffness was extracted from a standardized, manually defined region of interest at the brain vertex. A previously validated 3D MRE pattern-analysis support vector machine (SVM) classifier was applied to this prospective cohort as a secondary exploratory comparator. Group differences were assessed using Wilcoxon rank-sum tests. Diagnostic performance was evaluated using receiver operating characteristic analysis as an exploratory assessment of discrimination.
Results:
Median stiffness was higher in iNPH-group than CI-group (1.62 vs 0.98 kPa; p < 0.001). Stiffness demonstrated excellent discriminatory ability, yielding an apparent AUC of 1.00 with an optimal threshold of 1.43 kPa. The 3D SVM decision metric also yielded an apparent AUC of 1.00. The pretrained SVM model classified 18 of 20 subjects correctly (accuracy: 90.0% [95% CI: 68.3%-98.8%]). While promising, these performance estimates should be interpreted cautiously given the limited sample size.
Conclusion:
This study provides preliminary evidence that a 2D MRE protocol may help discriminate iNPH from other disorders causing cognitive impairment, providing a foundation for larger multi-center validation studies for assessing reproducibility, generalizability, and potential clinical utility.
