Magnetic resonance imaging (MRI) is widely used in medicine to detect, diagnose, and treat diseases such as cancer, with image interpretation performed by experts. Quantitative MRI, which uses numerical measurements during the scans, may now offer greater accuracy, repeatability, and speed – but a new study requires strict quality control to reach its full potential.

Researchers from the National Institute of Standards and Technology (NIST) led the study from 11 institutions, which compared measurements from 27 MRI scanners from three vendors in nine clinical sites across the country. To provide benchmarks and untangle sources of distortion and variation, the study used a tissue proxy, or “phantom,” originally developed at NIST to evaluate the performance of MRI machines and related software.

MRI uses magnetic fields and radio waves to make the internal structures of the body, especially soft tissues, visible. Traditional MRI poses several challenges. In addition to subjective image analysis, the performance of the MRI scanner may vary or different devices may produce conflicting images of the same patient.

Quantitative MRI offers prospects for more consistent disease detection, diagnosis, and treatment without the need for a tissue biopsy. Ideally, numerical measurements of tumors and other disease markers could be reproduced over time for many different patients, scanners, and clinics – potentially reducing medical costs. Organizations like the International Society for Magnetic Resonance in Medicine promote quantitative MRI, but it’s currently only used in research and not in routine patient care.

The new study compared MRI scanner readings of a value called T1, a property of water molecules that can depend on surrounding tissue. T1 of the pixels in images is one of the parameters that clinical MRI systems could measure. In contrast, subjective interpretations of MRI images consider “T1-weighted” judgments that are qualitative rather than numerical.

“One big difference between the T1-weighted and the current T1-weighted is that the T1-weighted is not quantitative,” said study leader Katy Keenan from NIST. “The pixel values ​​themselves cannot be compared with pixel values ​​in other data sets. It is not easy to compare T1-weighted data across data sets. With carefully recorded T1, a comparison of the pixel values ​​is possible because they have a quantitative meaning. “

The study found that T1 measurements can be subject to significant biases and variations. There was no uniform discrepancy pattern between the providers, so that a diagnostic threshold value determined on one MRI system cannot be transferred to other MRI systems. In some cases, these variations could make a clinical difference in diagnosing a benign versus a malignant brain tumor and seriously affect patient care, the study said.

To remedy this, the study recommended the introduction of rigorous quality control procedures for quantitative MRI to promote confidence and stability in measurement techniques and to transfer measurement thresholds for diagnosis, disease progression, and treatment monitoring from research institutions to the entire clinical community. The study results reflect earlier findings from other researchers.

Previous attempts to use T1 values ​​to categorize brain tumors have been hampered by technical inconsistencies and found to be unreliable. However, recent advances in quantitative measurement methods have resulted in improvements in accuracy, repeatability, and acquisition speed. The new study is a step towards the application of diagnostic threshold T1 measurement at multiple clinical sites.


Presentation: KE Keenan, Z. Gimbutas, A. Dienstfrey, KF Stupic, MA Boss, SE Russek, TL Chenevert, PV Prasad, J. Guo, WE Reddick, KM Cecil, A. Shukla-Dave, DA Nunez, AS Konar, M. Liu, SR Jambawalikar, L. Schwartz, J. Zheng, P. Hu, and EF Jackson. Multi-site, multi-platform comparison of the MRI T1 measurement with the system phantom. PLUS ONE . 06/30/2021.


Please enter your comment!
Please enter your name here