Sora Neuroscience, Inc. · Class II · Cleared Jun 6, 2025
| K-number | K251009 |
| Device name | Cirrus Resting State fMRI Software |
| Applicant | Sora Neuroscience, Inc. |
| Product code | QIH |
| Device class | Class II |
| Decision date | Jun 6, 2025 |
| Decision | Substantially Equivalent |
| Regulation | 892.2050 |
Cirrus Resting State fMRI Software is a medical imaging post-processing application that analyzes resting state functional MRI (BOLD fMRI) data from brain scans. It generates task-analogous correlation maps for three brain networks (sensorimotor, language, and vision) in DICOM format for visualization by surgeons, radiologists, and clinicians. The software is adjunctive and not intended to replace direct functional mapping procedures.
Both Cirrus and its predicate device perform resting state BOLD fMRI post-processing using the same scientific principle: spontaneous neural activations in resting state data exhibit temporal correlation within functionally connected brain regions. Both co-register BOLD fMRI with anatomical scans, perform fMRI preprocessing, compute correlation matrices, and generate task-analogous motor, language, and vision resting state maps. Technological differences in preprocessing pipelines and analysis methodologies do not raise new safety or effectiveness questions.
ISO 13485:2016 (quality management), IEC-62304:2006 + A1:2015 (software life cycle), ANSI/AAMI/ISO 14971:2019 (risk management), AAMI TIR57:2016/R2019 (cybersecurity), and NEMA PS 3.1-3.20 2024e (DICOM). Software underwent unit testing, verification testing, clinical performance validation, host platform testing, and cybersecurity penetration testing.
Cirrus shares the same indications for use and underlying scientific principle as the predicate Quicktome Software Suite. Clinical validation demonstrated Cirrus sensorimotor and language maps had mean spatial correlation of 0.826 and 0.793 respectively with published literature references, and vision maps achieved ROC area of 0.84 compared to task-activated vision maps. Validation across multiple patient populations (ages 3-71, brain tumor and epilepsy patients) and scanner manufacturers (Siemens and GE) at 99.9% confidence demonstrated comparability, establishing that technological differences do not introduce new safety or effectiveness concerns.
View the full FDA submission: accessdata.fda.gov