Comparison of quality control methods for automated diffusion tensor imaging analysis pipelines

January 2020 
Journal: PLOS One
Lead Author: Seyyed M. H. Mohammed

All Authors: Seyyed M. H. Haddad, Christopher J. M. Scott, Miracle Ozzoude, Melissa F. Holmes, Stephen R. Arnott, Nuwan D. Nanayakkara, Joel Ramirez, Sandra E. Black, Dar Dowlatshahi, Stephen C. Strother, Richard H. Swartz, Sean Symons, Manuel Montero-Odasso, ONDRI Investigators, Robert Bartha

Large research studies, run out of different labs/locations, can utilize different technologies (e.g. branded equipment). When brain imaging is involved, the output/records, results in huge data files. The processing of these files, to standardize them for analysis, requires extensive programming. To be practical and ensure quality control (QC), this programming takes the form of fully automatic pipelines/algorithms. In this study, three automatic processing pipelines were tested, using brain images from ONDRI participants. A preferred method was selected.