NCSI: Multiplex Network Analysis for Multimodal Neuroimaging

Status	Completed
Mentor Name	Konasale Prasad
Mentor's XSEDE Affiliation	Research Allocation
Mentor Has Been in XSEDE Community	1-2 years
Project Title	Multiplex Network Analysis for Multimodal Neuroimaging
Summary	We are developing a high-throughput data pipeline for generating multiplex networks from T1W, diffusion-weighted MRI, and fMRI imaging modalities. A multiplex network represents these diverse modes of MRI data as one graph-theoretic structure. Students will assist in the deployment of this data pipeline to the HPC context. This includes configuration of complex computing environments, code performance evaluation, and developing an understanding of image processing concepts.
Job Description	Students will be expected to work with bash command line interface, and multiple APIs for software suites associated with brain MRI data processing. During pipeline deployment phases of the study, students will learn while helping configure computing environments, and re-factor existing code where necessary to integrate the pipeline into the HPC environment. During piloting phases of the study, students will assist in code performance assessments, including quantifying resource use requirements of pipeline steps, and assessing quality of outputs. Students will help prepared summary documents of project progress.
Computational Resources	This project uses the Pittsburgh Super-computing Center's (PSC) Bridges-2 system regular memory nodes. This project relies on large sample size neuroimaging data.
Contribution to Community	We operate out of Pittsburgh's Western Psychiatric Institute and Clinic (WPIC), a non-profit mental health service provider and in-patient behavioral health hospital, and a leading neuroscience research center that investigates neuroscience of psychiatric disorders. This project supports ongoing brain imaging research at WPIC, administered through Pitt and the UPMC hospital system. Results of this project are applicable to this behavioral health research, and psychosis research in particular, in terms of elucidating the neurobiological mechanisms underlying these disorders. Overall goal is to design novel treatment paradigms for psychiatric disorders.
Position Type	Apprentice
Training Plan	Students will be given an orientation to project background, goals, and HPC fundamentals. Appropriate to their skill level, student will then be tasked with various aspects of pipeline configuration. At the most basic level, students will be trained to evaluate contents of files systems (i.e. tally file outputs) to track success of pipeline steps; next they will be tasked with automating this quality checking process. As students advance in their training they will be given the opportunity to work on primarily pipeline processes including working with HPC environments interactively for development purposes, refactoring existing code that may not be HPC-ready, committing changes to GitHub repositories, and submitting data processing jobs.
Student Prerequisites/Conditions/Qualifications	• Knowledge of Linux file systems and the bash command line interface required. • 1+ years of software and code development experience required. • 1+ years MATLAB experience required. • Github experience preferred. • SLURM experience a benefit but not required. • Singularity experience a benefit but not required. • Previous HPC experience desired. • Students will benefit from having some background in either MRI physics, signal processing, or neuroscience and psychiatry but not required.
Duration	Semester
Start Date	08/01/2021
End Date	12/20/2021