Chest-CAD
K210666Imagen Technologies, Inc · cleared 2021-07-20 · product code MYN · Radiology
Premarket evidence — what FDA accepted
Device typesamd
source quote (p.5)
“Chest-CAD is a computer-assisted detection (CADe) software device designed to assist physicians in identifying suspicious regions of interest (ROIs) in adult chest X-rays. The subject device is a software-only device.”
Algorithmdeep learning algorithms for computer vision; Artificial Neural Networks
source quote (p.5)
“Chest-CAD detects suspicious ROIs by analyzing radiographs using deep learning algorithms for computer vision and provides relevant annotations to assist physicians with their interpretations. Artificial Neural Networks”
Adaptive (vs locked)FDA source did not state this
PCCPFDA source did not state this
Cybersecurity addressedFDA source did not state this
Validation studies (2)
Bench
n=20,000 cases · 12 site(s)
endpoints: sensitivity; specificity; AUC
Retrospective clinical
n=238 cases · 9 site(s)
endpoints: AUC; sensitivity; specificity
Reported performance (9 observations)
sensitivity0.908CI 0.905, 0.911
source quote (p.8)
“The results of the standalone testing demonstrated that Chest-CAD detects suspicious ROIs with high sensitivity (0.908; 95% Wilson's Confidence Interval: 0.905, 0.911)”
specificity0.887CI 0.885, 0.889
source quote (p.8)
“high specificity (0.887; 95% Wilson's Confidence Interval: 0.885, 0.889)”
aurocas written: “auc”0.976CI 0.975, 0.976
source quote (p.8)
“high Area Under the Curve (AUC) of the Receiver Operating Characteristic (ROC) curve (0.976, 95% Bootstrap Confidence Interval: 0.975, 0.976).”
aurocas written: “Reader AUC (Aided)”0.894CI 0.879, 0.909
source quote (p.11)
“reader AUC estimates improved from 0.836 (95% Bootstrap CI: 0.816, 0.856) to 0.894 (95% Bootstrap CI: 0.879, 0.909).”
aurocas written: “Reader AUC (Unaided)”0.836CI 0.816, 0.856
source quote (p.11)
“reader AUC estimates improved from 0.836 (95% Bootstrap CI: 0.816, 0.856) to 0.894 (95% Bootstrap CI: 0.879, 0.909).”
sensitivityas written: “Reader sensitivity (Aided)”0.856CI 0.850, 0.862
source quote (p.11)
“Reader sensitivity improved from 0.757 (95% Wilson's CI: 0.750, 0.764) to 0.856 (95% Wilson's CI: 0.850, 0.862).”
sensitivityas written: “Reader sensitivity (Unaided)”0.757CI 0.750, 0.764
source quote (p.11)
“Reader sensitivity improved from 0.757 (95% Wilson's CI: 0.750, 0.764) to 0.856 (95% Wilson's CI: 0.850, 0.862).”
specificityas written: “Reader specificity (Aided)”0.87CI 0.866, 0.873
source quote (p.11)
“Reader specificity improved from 0.843 (95% Wilson's CI: 0.839, 0.847) to 0.870 (95% Wilson's CI: 0.866, 0.873).”
specificityas written: “Reader specificity (Unaided)”0.843CI 0.839, 0.847
source quote (p.11)
“Reader specificity improved from 0.843 (95% Wilson's CI: 0.839, 0.847) to 0.870 (95% Wilson's CI: 0.866, 0.873).”
Each value carries its own analysis unit and task — never compare or pool across devices. Source: 510(k) summary PDF.
Predicate network
Postmarket — what happened after clearance
0
recalls in product code, 24mo
0
MAUDE reports in code, 12mo
-100%
vs code's own 3-yr baseline
0
drift signals on this device
Recall and MAUDE counts are product-code-level (reports aren't reliably attributable to one device). Signals are descriptive observables with sources — never a judgment that the device is unsafe or drifting. Snapshot 2026-07-08.
Reimbursement — how devices like this got paid
Not yet tracked — no payment pathway indexed for this clearance (the reimbursement corpus is a growing seed set).