Premarket evidence — what FDA accepted

Device typesamd

source quote (p.4)

“iQ-solutions™ is a software medical device intended for automatic annotation, visualization, and quantification of segmentable brain structures from a set of brain MRI scans.”

Algorithmpre-trained convolutional neural networks (CNN)

source quote (p.6)

“iQ-solutions™ analysis module consists of pre-trained convolutional neural networks (CNN) that have been verified and validated to segment the specific brain structures and create binary masks accordingly using the incoming head MRI scans.”

Adaptive (vs locked)FDA source did not state this

PCCPFDA source did not state this

Cybersecurity addressedFDA source did not state this

Validation studies (3)

Bench

n=2,159 patients

endpoints: For classification modules, accuracy is higher than 0.9 to be accepted; for segmentation modules, DICE score is higher than 0.6 (for brains the DICE should be higher than 0.9) or R² is higher than 0.7 to be accepted; For modules that are lack of ground truths (scaling factor estimation, lesion inpainting, and substructure volume change estimation), the metrics are better than previously accepted results to be accepted; For all models, the manual acceptable rate should be higher than 80% to call it acceptable

standards: ISO 13485:2016 Medical devices - Quality management systems - Requirements for regulatory purposes, ISO 14971:2019 Medical devices – Application of risk management to medical devices, IEC 62304:2006 + A1:2015 Medical device software - software life cycle processes, IEC 62366-1:2015 Medical devices - Part 1: Application of usability engineering to medical devices, ISO 12052:2006 Digital Imaging and communication in Medicine (DICOM), CFR 21 Part 820 Quality System Regulation for Medical Devices

Bench

n=1,570 patients

endpoints: For classification modules, accuracy is higher than 0.9 to be accepted; for segmentation modules, DICE score is higher than 0.6 (for brains the DICE should be higher than 0.9) or R² is higher than 0.7 to be accepted; For modules that are lack of ground truths (scaling factor estimation, lesion inpainting, and substructure volume change estimation), the metrics are better than previously accepted results to be accepted; For all models, the manual acceptable rate should be higher than 80% to call it acceptable

standards: ISO 13485:2016 Medical devices - Quality management systems - Requirements for regulatory purposes, ISO 14971:2019 Medical devices – Application of risk management to medical devices, IEC 62304:2006 + A1:2015 Medical device software - software life cycle processes, IEC 62366-1:2015 Medical devices - Part 1: Application of usability engineering to medical devices, ISO 12052:2006 Digital Imaging and communication in Medicine (DICOM), CFR 21 Part 820 Quality System Regulation for Medical Devices

Bench

n=81 cases

endpoints: For classification modules, accuracy is higher than 0.9 to be accepted; for segmentation modules, DICE score is higher than 0.6 (for brains the DICE should be higher than 0.9) or R² is higher than 0.7 to be accepted; For modules that are lack of ground truths (scaling factor estimation, lesion inpainting, and substructure volume change estimation), the metrics are better than previously accepted results to be accepted; For all models, the manual acceptable rate should be higher than 80% to call it acceptable

standards: ISO 13485:2016 Medical devices - Quality management systems - Requirements for regulatory purposes, ISO 14971:2019 Medical devices – Application of risk management to medical devices, IEC 62304:2006 + A1:2015 Medical device software - software life cycle processes, IEC 62366-1:2015 Medical devices - Part 1: Application of usability engineering to medical devices, ISO 12052:2006 Digital Imaging and communication in Medicine (DICOM), CFR 21 Part 820 Quality System Regulation for Medical Devices

Reported performance (5 observations)

accuracyas written: “Accuracy (Sequence Classification)”100

source quote (p.15)

“Accuracy = 100%”

diceas written: “DICE (Brain Extraction)”0.982

source quote (p.15)

“DICE = 0.982”

diceas written: “DICE (White matter hyperintensity Segmentation)”0.789

source quote (p.15)

“DICE = 0.789”

diceas written: “DICE (Contrast-Enhancing Lesion Segmentation)”0.79

source quote (p.15)

“DICE = 0.790”

diceas written: “DICE (Brain Tissue Segmentation)”0.972

source quote (p.15)

“DICE = 0.972”

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

3

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).