Multi-dimensional MR fingerprinting for combined relaxometry-diffusion parameter mapping of the breast

Motivation: This work was driven by the need for efficient, high-fidelity multi-parametric MRI for advanced breast cancer applications, such as treatment response prediction. Goal(s): To develop a streamlined single-scan approach for acquiring co-aligned relaxometry and diffusion parametric maps of the breast, addressing challenges in multi-contrast MRI, including acquisition speed, image distortion, alignment and post-processing. Approach: We developed a multi-dimensional breast MR fingerprinting (mdMRF) sequence, incorporating multiple T1-, T2- and diffusion-preparation modules, with an acquisition time of 23 seconds/slice. Results: Our mdMRF T1, T2 and apparent diffusion coefficient (ADC) values agreed with reference methods in both phantom validations and in-vivo experiments, particularly in cancer patients. Impact: We present the first combined relaxometry-diffusion MR fingerprinting framework for breast imaging, paving the way for advancements in predictive breast cancer imaging, including improved capabilities for prediction of treatment response.