Diagnostic performance of a new multicontrast one‐minute full brain exam (EPIMix) in neuroradiology: A prospective study. Issue 6 (1st April 2019)

• Record Type: Journal Article
• Authors: Delgado, Anna F.; Kits, Annika; Bystam, Jessica; Kaijser, Magnus; Skorpil, Mikael; Sprenger, Tim; Skare, Stefan
• Is Part Of: Journal of magnetic resonance imaging. Volume 50:Issue 6(2019)
• Page Start: 1824
• Page End: 1833
• Publication Date: 2019-04-01
• Subjects: EPIMix -- neuroradiology -- brain -- diagnostic performance -- minute -- MRI; Magnetic resonance imaging -- Periodicals; 616
• DOI: 10.1002/jmri.26742
• Languages: English
• ISSNs: 1053-1807

Abstract

Abstract : Background: Clinical MRI protocols are time‐consuming; hence, new faster techniques are needed. One new fast multicontrast MRI technique, called echo planar image mix (EPIMix) (including contrasts T1 ‐FLAIR, T2 ‐weighted, diffusion‐weighted images [DWI], apparent diffusion coefficient [ADC], T2 *‐weighted, and T2 ‐FLAIR images) needs to be tested. Purpose: To assess if EPIMix has comparable diagnostic performance as routine clinical brain MRI. Study Type: Prospective. Population: A consecutive series of 103 patients' brain MRI (January 2018 to May 2018). Field Strength/Sequence: 1.5 T or 3T. EPIMix and routine clinical protocol (clinical MRI included all or some of the contrasts T1 ‐FLAIR, T2 ‐weighted, DWI, T2 *‐weighted, T2 ‐FLAIR, 3D‐FSE). Assessment: Two neuroradiologists assessed EPIMix and clinical scans and categorized the images as abnormal or normal and described diagnosis, artifacts, diagnostic confidence image quality, and comparison of imaging time. Statistical Tests: Pivot tables with diagnostic performance calculated by receiver operating characteristics (ROC) and the area under curve (AUC). Disease categorization and image quality measures were evaluated. The study protocol is published at ClinicalTrials.gov NCT03338270. Results: After exclusion of 21 patients, 82 patients had a routine clinical MRI with comparable contrasts to EPIMix and were evaluated. The diagnostic performance to categorize a full brain MRI investigation as abnormal or normal was comparable between EPIMix (AUC 0.99 (95% confidence interval [CI] 0.97–1.00) and 0.99 (95% CI 0.97–1.00)) and routine clinical MRI ( n = 82). Sensitivity was 95% (95% CI 88–95) and 93% (95% CI 86–98), and specificity 100% (95% CI 97–100) and 100% (95% CI 90–100). Disease categorization was congruent between EPIMix and clinical routine MRI in 90% (reader 2) and 93% (reader 1). Image quality was generally rated lower for EPIMix ( P < 0.001). Imaging time was 78 seconds for EPIMix and for the same contrasts 12 minutes 29 seconds for conventional 3T MRI. Data Conclusion: EPIMix has comparable diagnostic performance (disease identification and categorization) for most patients investigated in clinical routine. Level of Evidence 2 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2019;50:1824–1833.