Case Study: Stanford University achieves faster, scalable personalized brain-stimulation workflows with Flywheel

Automating Workflows in Stanford's Brain Stimulation Lab

Stanford University’s Brain Stimulation Lab, developing personalized transcranial magnetic stimulation (Stanford Neuromodulation Therapy) for treatment‑resistant depression, faced a labor‑intensive imaging and targeting workflow that slowed patient treatment and made provenance and scaling difficult. To meet growing multi‑site studies and faster turnaround needs, the team adopted the Flywheel platform — including Flywheel Gears and the Flywheel SDK — to manage scans, processing, and automated analysis.

Using Flywheel, the lab containerized and daisy‑chained processing steps (reconstruction, QA, fMRIPrep, and a custom targeting algorithm) with two custom Gears and rule‑based execution, plus cloud‑scalable compute and built‑in provenance/versioning. Flywheel cut typical pre‑treatment turnaround from about 1–2 days to 8–12 hours (and can accelerate inpatient prep to under 12 hours), enabled full reconstruction and BIDS curation within an hour or two of scanning, and freed staff time while improving reproducibility and workflow tracking.

Stanford University

Azeezat Azeez

Postdoctoral Scholar