Technical Note: Human tissue‐equivalent MRI phantom preparation for 3 and 7 Tesla. Issue 8 (17th July 2021)

• Record Type: Journal Article
• Authors: Woletz, Michael; Roat, Sigrun; Hummer, Allan; Tik, Martin; Windischberger, Christian
• Is Part Of: Medical physics. Volume 48:Issue 8(2021)
• Page Start: 4387
• Page End: 4394
• Publication Date: 2021-07-17
• Subjects: high field -- MRI phantom -- phantoms – physical -- quality control -- quantitative imaging/analysis -- relaxivity -- T1 -- T2; Medical physics -- Periodicals; Medical physics; Geneeskunde; Natuurkunde; Toepassingen; Biophysics; Periodicals; Periodicals; Electronic journals; 610.153
• DOI: 10.1002/mp.14986
• Languages: English
• ISSNs: 0094-2405

Abstract

Abstract : Purpose: While test objects (phantoms) in magnetic resonance imaging (MRI) are crucial for sequence development, protocol validation, and quality control, studies on the preparation of phantoms have been scarce, particularly at fields exceeding 3 Tesla. Here, we present a framework for the preparation of phantoms with well‐defined T1 and T2 times at 3 and 7 Tesla. Methods: Phantoms with varying concentrations of agarose and Gd‐DTPA were prepared and measured at 3 and 7 Tesla using T1 and T2 mapping techniques. An empirical, polynomial model was constructed that best represents the data at both field strengths, enabling the preparation of new phantoms with specified combinations of both T1 and T2 . Instructions for three different tissue types (brain gray matter, brain white matter, and renal cortex) are presented and validated. Results: T1 times in the samples ranged from 698 to 2820 ms and from 695 to 2906 ms, whereas T2 times ranged from 39 to 227 ms and from 34 to 235 ms for 3 and 7 Tesla scans, respectively. Models for both relaxation times used six parameters to represent the data with an adjusted R² of 0.998 and 0.997 for T1 and T2, respectively. Conclusion: Based on the equations derived from the current study, it is now possible to obtain accurate weight specifications for a test object with desired T1 and T2 relaxation times. This will spare researchers the laborious task of trail‐and‐error approaches in test object preparation attempts.