Methinks NCCT Stroke

K250685

Methinks Software S.L. · cleared 2025-06-16 · product code QAS · Radiology

Premarket evidence — what FDA accepted

Device typesamd

source quote (p.5)

“Methinks NCCT Stroke is a radiological computer-assisted triage and notification software device. The Methinks NCCT Stroke device is an AI/ML Software as a Medical Device.”

Algorithmartificial intelligence algorithm

source quote (p.4)

“Methinks NCCT Stroke uses an artificial intelligence algorithm to analyze images and highlight cases with suspected (1) ICH and (2) LVO in the cloud in parallel to the ongoing standard of care image interpretation.”

Adaptive (vs locked)FDA source did not state this

PCCPFDA source did not state this

Cybersecurity addressedYes

source quote (p.8)

“Cybersecurity has been considered and addressed as part of the software verification process, in line with Section 524B of the FD&C Act as a Cyber device. Methinks implemented a risk-based cybersecurity strategy including secure design principles, vulnerability assessment, Software Bill of Materials (SBOM), and penetration testing. These activities ensure the software's safety, effectiveness, and robustness against potential cybersecurity threats.”

Validation studies (2)

Retrospective clinical

n=358 cases

endpoints: evaluate the software's performance in identifying NCCT head images containing intracranial hemorrhage (ICH); evaluate the software's performance in identifying Large Vessel Occlusion (LVO) findings

standards: EN ISO 14971:2019, ISO 62304:2015

Reader study (MRMC)

n=335 cases

endpoints: expert non-inferiority for sensitivity; non-expert superiority for sensitivity

Reported performance (7 observations)

sensitivity0.947CI 89.3% - 97.8%

source quote (p.9)

“Specifically, ICH performance was observed at Se: 94.7% (95% CI: 89.3% - 97.8%)”

specificity0.995CI 97.5% - 99.9%

source quote (p.9)

“Sp: 99.5% (95% CI: 97.5% - 99.9%)”

sensitivityas written: “LVO sensitivity”0.764CI 67.3% - 83.9%

source quote (p.9)

“and LVO was observed at Se: 76.4% (95% CI: 67.3% - 83.9%)”

specificityas written: “LVO specificity”0.911CI 86.6% - 94.5%

source quote (p.9)

“and Sp: 91.1% (95% CI: 86.6% - 94.5%).”

sensitivityas written: “LVO sensitivity (Se) for expert neuroradiologists”0.5CI 40.1% - 59.9%

source quote (p.9)

“LVO sensitivity (Se) for expert neuroradiologists was 50.0% (40.1% - 59.9%)”

time_to_resultas written: “Time to notification of NCCT-ICH (minutes) - Mean”1.43CI 1.36 - 1.50

source quote (p.9)

“Time to notification of NCCT-ICH (minutes) 1.43 1.36 1.50”

time_to_resultas written: “Time to notification of NCCT-LVO (minutes) - Mean”1.42CI 1.36 - 1.48

source quote (p.9)

“Time to notification of NCCT-LVO (minutes) 1.42 1.36 1.48”

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

recalls in product code, 24mo

MAUDE reports in code, 12mo

—

vs code's own 3-yr baseline

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).