NeuroICH
K241719Neurocareai Inc. · cleared 2024-11-07 · product code QAS · Radiology
Premarket evidence — what FDA accepted
Device typesamd
source quote (p.6)
“NeurolCH is a software-only parallel workflow tool designed for use by hospital networks and trained clinicians to identify and communicate prioritized images of specific patients to an appropriate specialist such as neurovascular or neurosurgical specialist independent of the standard of care workflow. The device uses an artificial intelligence algorithm to analyze non-contrast CT images of the head acquired in the acute setting for findings suggestive of intracranial hemorrhage (ICH) in parallel to the ongoing standard of care image interpretation and notify an appropriate clinician of these findings.”
Algorithmdeep learning artificial intelligence algorithm
source quote (p.6)
“The standalone software device automatically receives and analyzes non-contrast head CT (NCCT) studies of patients undergoing stroke protocol, for image features that indicate the presence of an intracranial hemorrhage (ICH) using deep learning artificial intelligence algorithm, and upon detection of a suspected ICH case, sends a notification along with non-diagnostic image on mobile application to alert a specialist clinician.”
Adaptive (vs locked)No
PCCPNo
Cybersecurity addressedNo
Validation studies (1)
Retrospective clinical
n=376 cases
endpoints: Sensitivity; Specificity; Accuracy; Area Under the Curve (AUC); Time-to-Notification (TTN)
Reported performance (5 observations)
sensitivity94.81CI 89.68% - 97.43%
source quote (p.9)
“Sensitivity and specificity on the primary dataset were observed to be 94.81% (89.68% - 97.43%) and 92.53% (88.50% - 95.21%), respectively.”
specificity92.53CI 88.50% - 95.21%
source quote (p.9)
“Sensitivity and specificity on the primary dataset were observed to be 94.81% (89.68% - 97.43%) and 92.53% (88.50% - 95.21%), respectively.”
aurocas written: “auc”0.9367
source quote (p.9)
“In addition, the accuracy and area under the receiver operating characteristic curve (AUC) were 93.35% (90.37% - 95.45%) and 0.9367 respectively, demonstrating the clinical utility and potential benefits of the classifier based on the imaging study results.”
accuracyas written: “Accuracy”93.35CI 90.37% - 95.45%
source quote (p.9)
“In addition, the accuracy and area under the receiver operating characteristic curve (AUC) were 93.35% (90.37% - 95.45%) and 0.9367 respectively, demonstrating the clinical utility and potential benefits of the classifier based on the imaging study results.”
time_to_resultas written: “Time to Notification (TTN)”0.37CI ± 0.20 minutes
source quote (p.15)
“The average time to alert a specialist by NeuroICH was 0.37 ± 0.20 minutes, which is lower than the average time to open an exam seen in the Standard of Care 18.3±14.2 minutes and comparable to the time reported by the predicate device Viz ICH 0.49±0.08 minutes.”
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
—
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).