Purpose: In this work, we present the MASSIVE (Multiple Acquisitions for Standardization of Structural Imaging Validation and Evaluation) brain dataset of a single healthy subject, which is intended to facilitate diffusion MRI (dMRI) modeling and methodology development.
Methods: MRI data of one healthy subject (female, 25 years) were acquired on a clinical 3 Tesla system (Philips Achieva) with an eight-channel head coil. In total, the subject was scanned on 18 different occasions with a total acquisition time of 22.5 h. The dMRI data were acquired with an isotropic resolution of 2.5 mm3 and distributed over five shells with b-values up to 4000 s/mm2 and two Cartesian grids with b-values up to 9000 s/mm2 .
Results: The final dataset consists of 8000 dMRI volumes, corresponding B0 field maps and noise maps for subsets of the dMRI scans, and ten three-dimensional FLAIR, T1 -, and T2 -weighted scans. The average signal-to-noise-ratio of the non-diffusion-weighted images was roughly 35.
Conclusion: This unique set of in vivo MRI data will provide a robust framework to evaluate novel diffusion processing techniques and to reliably compare different approaches for diffusion modeling. The MASSIVE dataset is made publically available (both unprocessed and processed) on www.massive-data.org. Magn Reson Med 77:1797-1809, 2017. © 2016 International Society for Magnetic Resonance in Medicine.
Keywords: brain dataset; diffusion MRI; evaluation; methods development; modeling; structural MRI.
© 2016 International Society for Magnetic Resonance in Medicine.