Premium
Reference range determination for imaging biomarkers: Myocardial T 1
Author(s) -
Higgins David M.,
Keeble Claire,
Juli Christoph,
Dawson Dana K.,
Waterton John C.
Publication year - 2019
Publication title -
journal of magnetic resonance imaging
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.563
H-Index - 160
eISSN - 1522-2586
pISSN - 1053-1807
DOI - 10.1002/jmri.26683
Subject(s) - reference range , confidence interval , medicine , nuclear medicine , range (aeronautics) , statistics , population , reference values , undersampling , mathematics , computer science , artificial intelligence , materials science , environmental health , composite material
Background Imaging biomarkers, such as the T 1 relaxation time of the myocardium using MRI, can be valuable in cardiac medicine if they are properly validated. Consensus statements recommend that for myocardial T 1 , each investigator should establish a reference range. Purpose To describe a statistically valid method for determining and reporting the reference range in each center, which simultaneously minimizes the twin risks of undersampling, leading to a uselessly uncertain range, and oversampling, which exposes volunteers to unnecessary scanning and wastes resources. Study Type Cohort. Population In all, 278 normal human subjects without cardiac disease from two cardiac MR centers. Field Strength/Sequence 1.5 T and 3 T; Modified Look–Locker Inversion recovery sequence. Assessment The T 1 relaxation time was estimated from multiple samples of tissue magnetization after inversion. A valid method for calculating a reference range was used. Statistical Tests Shapiro–Wilk test for normality; Tukey robust approach for identification of outliers; reference range calculation with confidence intervals. Results Reference ranges for measurement of myocardial T 1 were calculated, with confidence intervals, enabling comparison with clinically important differences. At 3 T: 1129 to 1301 msec at site 1 ( n = 21) and 1160 to 1309 msec at site 2 ( n = 59), and at 1.5 T at site 2: 933 to 1020 msec (male, n = 130) and 965 to 1054 msec (female, n = 68). The 3 T reference range from site 1 was successfully benchmarked against the 3 T reference range at site 2. Data Conclusion Myocardial T 1 reference ranges can be properly characterized, enabling clinical comparison to a valid reference range with known confidence intervals, using methodology similar to that described in this report. Level of Evidence: 3 Technical Efficacy Stage: 3 J. Magn. Reson. Imaging 2019;50:771–778.