| K-number | K242467 |
| Device name | IQ-UIP |
| Applicant | Imbio, Inc. |
| Product code | QWO |
| Device class | Class II |
| Decision date | Dec 19, 2024 |
| Decision | Substantially Equivalent |
| Regulation | 892.2085 |
IQ-UIP is computer-aided software that analyzes chest CT scans to notify specialists at interstitial lung disease (ILD) centers of radiological findings suggestive of usual interstitial pneumonia (UIP) pattern. It uses artificial intelligence to identify positive findings on a separate worklist application parallel to standard radiological interpretation and provides informational summary reports with clinical guideline references. The device does not alter original images and is not intended as a diagnostic device; clinicians remain responsible for final clinical determinations.
IQ-UIP is a software-only Class II device (product code QWO) under regulation 21 CFR 892.2085, identical to the predicate Fibresolve. Both use machine learning and artificial intelligence algorithms to analyze lung/chest CT images from adults over 22 years old. The key difference is that IQ-UIP displays referral reports on a dedicated worklist application separate from standard-of-care, whereas Fibresolve's viewing approach was unknown. Both devices do not alter original images and operate in parallel to standard workflow.
Not stated in this summary.
IQ-UIP demonstrates substantial equivalence to Fibresolve through shared fundamental characteristics: both are software-only AI/machine learning referral tools analyzing lung CTs for fibrotic lung disease in the same patient population (adults >22 years), using identical product codes and regulations, operating in parallel to standard care, and providing qualitative diagnostic information for clinician review rather than making final diagnoses. Performance data from 804 images with independent radiologist ground-truthing (AUC ROC 96.6, PPV 77.9) demonstrates comparable safety and effectiveness. The minor difference—IQ-UIP's dedicated worklist display versus Fibresolve's unknown viewing method—does not alter the fundamental intended use or clinical risk profile, as both require clinician review and final determination.
View the full FDA submission: accessdata.fda.gov