Flywheel

The Dana and David Dornsife Cognitive Neuroimaging Center (DNI) at the University of Southern California (USC) was facing challenges in distributing data to researchers across the university community. The data from scans conducted at the center had to be transferred to DVDs and delivered to individual researchers or pushed onto the researcher's own DICOM server. These manual processes often resulted in delays and quality concerns. Additionally, Dr. Jonas Kaplan, Assistant Research Professor of Psychology at USC's Brain and Creativity Institute and Co-Director of the DNI, was seeking a new platform to manage his lab's data. He wanted a solution that would allow him to access and curate his data quickly and accurately, while also providing a secure mechanism for sharing data with collaborators and students.

The University of Texas Medical Branch (UTMB) was facing challenges in realizing greater value from its imaging assets and collaborating more efficiently within and outside the system. The COVID-19 pandemic further amplified the need for efficient and remote collaboration on imaging research. UTMB wanted to leverage its imaging assets more fully as its radiology archive grew and it went live with digital pathology. The hospital aimed to create data sets and make them available for AI researchers. However, without a clear way to organize the process, this posed a significant challenge.

Neurologists treating patients with traumatic brain injury (TBI) have long faced a significant challenge: determining which patients with mild or moderate head injuries are at increased future risk of developing neurological problems such as dementia, mood disorders, and Parkinson’s disease, and which are not. Both in classification and outcome assessments, TBI scores are often exclusively symptom-based, and therefore too general to catch some brain injuries and prognoses. To improve the diagnosis, treatment and rehabilitation of patients with TBI, Dr. Geoffrey Manley, Vice Chairman of Neurological Surgery at the University of California, San Francisco, set up the Transforming Research and Clinical Knowledge in Traumatic Brain Injury (TRACKTBI) study 10 years ago. Today, 19 institutional partners in the TRACK-TBI NETWORK collect more than 3,000 data fields per subject, including outcome measures assessed at four time points post-injury: medical imaging, biospecimen samples, proteome test results and genomic information.

The Yassa Lab at the University of California, Irvine (UCI), led by Dr. Michael Yassa, was facing several challenges. They were struggling with managing multi-center collaboration involving the collection of large data sets, quality control, analysis, and submission to NIH databases. The growing data and analytic complexity were impeding data reuse and scientific reproducibility. They were also looking for ways to best support and collaborate with other labs in the UC Irvine community. The lab was involved in a multicenter collaboration studying biomarkers of Alzheimer's disease in Down syndrome, which required secure sharing and processing of a variety of data.

The Brain Imaging Data Generation and Education Center (BRIDGE) at Carnegie Mellon University (CMU) and the University of Pittsburgh (Pitt) has been an early adopter of BIDS' (Brain Imaging Data Structure). BIDS is an increasingly adopted standard of data organization that allows researchers to more easily share neuroimaging data and software tools across the broad range of research conducted by users scanning at their facilities. The BRIDGE Center leadership sees this technology for standardizing (i.e. organizing, annotating, and describing) data as an important facilitator for replicable analyses and advancing research collaboration to speed discovery. The need for efficient data practices became more evident when the Center decided to purchase another 3-Tesla MRI system in 2019, a decision that would greatly increase the amount of data acquired at the Center.

Stanford Medicine’s Brain Stimulation Lab is working on solutions for treatment-resistant depression, a condition that affects 5% of adults worldwide. The lab is studying the use of Repetitive Transcranial Magnetic Stimulation (rTMS), a therapy that involves activating or inhibiting the brain directly with electromagnetic fields. The lab's work is growing, and so is their need for smart data management. The lab originally used the Flywheel platform to store raw and reconstructed data and applied its basic tools for reconstruction and quality control. However, when they wanted to perform analysis, researchers were still downloading data to a static lab PC. This process was time-consuming and made it difficult to track data provenance.