The first widely useful standard for magnetic resonance imaging (MRI) of the breast has been developed by the National Institute of Standards and Technology (NIST).1,2
By developing a “phantom,” researchers were able to standardize MRIs of breast tissue and ensure quality control in comparing images within and between medical research studies. Phantoms mimic the response of human tissue to help test the performance of medical imaging systems.
NIST’s new breast phantom also can be used to compare MRI scanners and to train operators. The phantom supports quantitative MRI, which is increasingly used for breast cancer diagnosis, staging, and treatment monitoring as well as for imaging other parts of the body.
The prototype breast phantom was tested on MRI systems from 2 manufacturers in 3 configurations. It was able to produce accurate, quantitative images. It fits most MRI scanner designs and meets a full range of common clinical imaging needs. Its image data can be traced to international measurement standards.
The biggest design challenge was to create a lifelike mimic of both fat and fibroglandular tissue, said NIST project leader Kathryn Keenan, PhD. The design success led to industry’s rapid commercialization of the phantom, currently used by 5 research groups. The design of the phantom is flexible and modular. Its soft silicone shells can fit different MRI scanners, while its internal components are made of rigid polycarbonate to ensure regular geometry.
“I believe the NIST phantom can have a big impact, especially because we are seeing more use of quantitative MRI for breast cancer,” Keenan said. “It will be used by both researchers developing new techniques and by radiologists using techniques in the clinic.”
1. Keenan KE, Peski AP, Wilmes LJ, et al. Variability and bias assessment in breast ADC measurement across multiple systems. [Published online March 23, 2016] JMRI. doi:10.1002/jmri.25237.
2. Keenan KE, Wilmes LJ, Aliu SO, et al. Design of a breast phantom for quantitative MRI. [Published online March 7, 2016] JMRI. doi:10.1002/jmri.25214.