Prostate MR AI (VA10A)

K241770

Siemens Healthcare GmbH · cleared 2025-03-05 · product code QDQ · Radiology

Premarket evidence — what FDA accepted

Device typesamd

source quote (p.4)

“Prostate MR AI is a plug-in Radiological Computer Assisted Detection and Diagnosis Software device intended to be used”

AlgorithmArtificial intelligence algorithm trained on a database of prostate MR image series acquired according to the PI-RADS standard (non-contrast T2W and DWI image series), and corresponding radiological and/or biopsy findings.

source quote (p.10)

“Artificial intelligence algorithm trained on a database of prostate MR image series acquired according to the PI-RADS standard (non-contrast T2W and DWI image series), and corresponding radiological and/or biopsy findings.”

Adaptive (vs locked)FDA source did not state this

PCCPFDA source did not state this

Cybersecurity addressedFDA source did not state this

Validation studies (4)

Bench

n=222 images

endpoints: median of the Dice score between the AI algorithm results and the corresponding ground truth masks exceeds the threshold of 0.9; median of the normalized volume difference between the algorithm results and the corresponding ground truth masks is within a ±5% range; AI algorithm results as compared to any individual reader are statistically non-inferior based on variabilities that existed among the individual readers within the 5% margin of error and 5% significance level.

standards: ISO 14971 Third Edition 2019-12, IEC 62304 Edition 1.1 2015-06 CONSOLIDATED VERSION, IEC 82304-1 Edition 1.0 2016-10

Retrospective clinical

n=105 cases · 6 site(s)

endpoints: The case level sensitivity of the lesion detection is equal or greater than 0.80 for both radiology and pathology ground truth.; The false positive rate per case of the lesion detection is smaller than one false positive per case for radiology ground truth.; The accuracy of the PI-RADS classification of radiology ground truth lesions detected by the algorithm is equal or greater than 0.8.

standards: ISO 14971 Third Edition 2019-12, IEC 62304 Edition 1.1 2015-06 CONSOLIDATED VERSION, IEC 82304-1 Edition 1.0 2016-10

Retrospective clinical

n=115 cases · 6 site(s)

standards: ISO 14971 Third Edition 2019-12, IEC 62304 Edition 1.1 2015-06 CONSOLIDATED VERSION, IEC 82304-1 Edition 1.0 2016-10

Reader study (MRMC)

n=340 cases · 2 site(s)

endpoints: comparison of case-level diagnostic performance of aided and unaided reads using the reader-provided case-level LoS (RLoS)

standards: ISO 14971 Third Edition 2019-12, IEC 62304 Edition 1.1 2015-06 CONSOLIDATED VERSION, IEC 82304-1 Edition 1.0 2016-10

Reported performance (13 observations)

diceas written: “Median Dice score (segmentation)”0.9

source quote (p.15)

“The median of the Dice score between the AI algorithm results and the corresponding ground truth masks exceeds the threshold of 0.9”

sensitivityas written: “Case level sensitivity of lesion detection”0.8

source quote (p.15)

“The case level sensitivity of the lesion detection is equal or greater than 0.80 for both radiology and pathology ground truth.”

false_positive_rate_per_imageas written: “False positive rate per case of lesion detection”1

source quote (p.15)

“The false positive rate per case of the lesion detection is smaller than one false positive per case for radiology ground truth.”

accuracyas written: “Accuracy of PI-RADS classification”0.8

source quote (p.15)

“The accuracy of the PI-RADS classification of radiology ground truth lesions detected by the algorithm is equal or greater than 0.8.”

aurocas written: “AUROC (reader study, fully inclusive, unaided)”0.658

source quote (p.17)

“the average area under the ROC (Receiver Operating Characteristic) curve (AUROC) improved from 0.658 in unaided reading to 0.7010 in aided reading”

aurocas written: “AUROC (reader study, fully inclusive, aided)”0.701

source quote (p.17)

“the average area under the ROC (Receiver Operating Characteristic) curve (AUROC) improved from 0.658 in unaided reading to 0.7010 in aided reading”

aurocas written: “AUROC difference (reader study, fully inclusive, aided - unaided)”0.0252CI [0.001, 0.0493]

source quote (p.17)

“with a difference of 0.0252 (95% C.I. [0.001, 0.0493], P=0.040).”

aurocas written: “AUROC (reader study, maximally restrictive, unaided)”0.6579

source quote (p.17)

“In the maximally restrictive analysis, AUROC improved from 0.6579 in unaided reading to 0.6948 in aided reading”

aurocas written: “AUROC (reader study, maximally restrictive, aided)”0.6948

source quote (p.17)

“In the maximally restrictive analysis, AUROC improved from 0.6579 in unaided reading to 0.6948 in aided reading”

aurocas written: “AUROC difference (reader study, maximally restrictive, aided - unaided)”0.0368CI [0.0108, 0.0628]

source quote (p.17)

“with a difference of 0.0368 (95% C.I. [0.0108, 0.0628], P=0.006).”

agreement_kappaas written: “Fleiss' Kappa (reader study, fully inclusive, unaided)”0.283CI [0.242, 0.322]

source quote (p.18)

“In a supplemental analysis for the fully inclusive analysis scenario, Fleiss' Kappa for interreader agreement in per-case PI-RADS scores was 0.283 (95% C.I.: [0.242, 0.322]) for unaided reads”

agreement_kappaas written: “Fleiss' Kappa (reader study, fully inclusive, aided)”0.371CI [0.326, 0.411]

source quote (p.18)

“and 0.371 (95% C.I.: [0.326, 0.411]) for aided reads”

agreement_kappaas written: “Fleiss' Kappa difference (reader study, fully inclusive, aided - unaided)”0.087CI [0.051, 0.125]

source quote (p.18)

“with a difference of 0.087 (95% C.I. [0.051, 0.125]).”

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).