The transition to quantitative MRI requires accurate and traceable standards. Here, we provide an update on the NIST MRI Biomarker Measurement Service https://www.nist.gov/programs-projects/magnetic-resonance-imaging-mri-biomarker-measurement-service, which provides calibrations for MRI biomarkers including T1, T2, and diffusivity D. The calibration service is being extended to include biomarkers for non-gaussian diffusion, anisotropic diffusion, brain-MRS, and MSK related biomarkers. Associated with the calibration service is a medical phantom lending library https://www.nist.gov/programs-projects/nistnibib-medical-imaging-phantom-lending-library, cosponsored by NIH, that curates and provides easy access to calibrated phantoms. Here, we outline the challenges for rigorous definition and calibration of MRI biomarkers and provide case studies demonstrating their utility.
For most biomarkers a hierarchy of standards are needed. For T1, T2, and D, ideal samples that show single component mono-exponential behaviors and Gaussian diffusion are required. Environmental conditions, such as temperature, magnetic field, field noise, must be monitored and precisely measured during calibration. These reference samples, while not being representative of complex tissue, are needed to validate the accuracy and stability of the scanner and quantitative protocols. More complex biomimetic materials are then required to validate quantitative tissue measurements. These materials have complex multi-peak T1, T2 spectra, multi-exponential behavior, and non-gaussian diffusion. We present methods to obtain rigorous MRI-effective parameters that are of use to the MRI community but retain rigorous calibration and traceability. Finally, complex structures, such as realistic tumor mimics, are required to validate radiomic measurements where there is considerable heterogeneity. We discuss methods to provide a traceable chain from metrology NMR/MRI systems to clinical systems where many metrology requirements cannot be met.
The calibration of materials that go into traceable phantoms is only the first step in the quantification process. Phantom imaging conditions, e.g. temperature, must be accurately recorded and the phantoms must be monitored for stability. Accurate MRI readable thermometry and recalibration protocols are required.
We provide several case studies detailing the use of traceable MRI standards including monitoring scanners before and after major equipment failures and planned upgrades, use of quantitative phantoms to validate new and faster imaging protocols such as MR Fingerprinting, and the use of phantoms to assess non-Gaussian diffusion biomarkers for tumor characterization.