StrokeSENS ASPECTS Software Application

K250221

Circle Cardiovascular Imaging Inc. · cleared 2025-07-01 · product code POK · Radiology

Premarket evidence — what FDA accepted

Device typesamd

source quote (p.7)

“StrokeSENS ASPECTS is a stand-alone software device that uses machine learning algorithms to automatically process NCCT (non-contrast computed tomography) brain image data to provide an output ASPECTS score based on the Alberta Stroke Program Early CT Score (ASPECTS) guidelines.”

AlgorithmMachine learning algorithms (Deep Learning)

source quote (p.7)

“StrokeSENS ASPECTS is a stand-alone software device that uses machine learning algorithms to automatically process NCCT (non-contrast computed tomography) brain image data to provide an output ASPECTS score based on the Alberta Stroke Program Early CT Score (ASPECTS) guidelines. Technical Implementation: ML/AI (Deep Learning)”

Adaptive (vs locked)FDA source did not state this

PCCPFDA source did not state this

Cybersecurity addressedYes

source quote (p.13)

“StrokeSENS has been developed and tested to meet cybersecurity requirements using design vulnerability Assessments (Threat Models), SBOM's, NVD assessments, and Penetration Testing”

Validation studies (3)

Retrospective clinical

n=200 patients

endpoints: region-level Clustered ROC Analysis to demonstrate the standalone performance of the ASPECTS device with respect to the expert consensus reference standard

standards: ISO 13485:2016+A11:2021, IEC 62304:2006+A1:2015, IEC 62366:2015+A1:2020, ISO 14971:2019+A11:2021, NEMA 3.1-3.20 (2021)

Reader study (MRMC)

n=100 patients · 23 site(s)

endpoints: statistically significant improvement of 5.7% from 68.6% to 74.3% (p-value<0.001) in AUC for each reader with and without the support of the StrokeSENS ASPECTS tool

Bench

sample size not stated

endpoints: high Concordance Rate of 97.0% (i.e., the proportion of cases with a consensus score of Fair Concordance or above), with respect to the agreement between the device's auto-generated VCTA overlay and expert neuroradiologist assessment of ischemic tissue.

Reported performance (4 observations)

sensitivity0.706CI [69.2%, 72.1%]

source quote (p.13)

“a sensitivity of 70.6% [69.2%, 72.1%]”

specificity0.939CI [93.2%, 94.7%]

source quote (p.13)

“a specificity of 93.9% [93.2%, 94.7%]”

aurocas written: “auc”0.909CI [88.7%, 93.1%]

source quote (p.13)

“an AUC-ROC of 90.9% (95% CI = [88.7%, 93.1%])”

accuracy0.906CI [89.7%, 91.5%]

source quote (p.13)

“accuracy of 90.6 [89.7%, 91.5%]”

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).