HERA Z20 Diagnostic Ultrasound System; HERA Z20e Diagnostic Ultrasound System; HERA Z20s Diagnostic Ultrasound System; R20 Diagnostic Ultrasound System; HERA Z30 Diagnostic Ultrasound System; R30 Diagnostic Ultrasound System

K252018

Samsung Medison Co., Ltd. · cleared 2026-01-05 · product code IYN · Radiology

Premarket evidence — what FDA accepted

Device typehardware with ml
source quote (p.6)
“mobile, software controlled, diagnostic ultrasound system. Their function is to acquire ultrasound data and to display the data as 2D mode, Color Doppler mode, Power Doppler (PD) mode, M mode, Pulsed Wave (PW) Doppler mode, Continuous Wave (CW) Doppler mode, Tissue Doppler Imaging (TDI) mode, Tissue Doppler Wave (TDW) mode, ElastoScan Mode, Combined modes, MV-Flow mode, Multi-Image mode(Dual, Quad), 3D/4D mode. [...] The proposed HERA Z20 Diagnostic Ultrasound System; HERA Z20e Diagnostic Ultrasound System; HERA Z20s Diagnostic Ultrasound System; R20 Diagnostic Ultrasound System; HERA Z30 Diagnostic Ultrasound System; R30 Diagnostic Ultrasound System have included new AI-ML based software features: AbdomenAssist, BladderAssist, QualityCheck, PelvicAssist.”
AlgorithmAI-ML based software features including AbdomenAssist, BladderAssist, QualityCheck, PelvicAssist, NerveTrack, EzVolume (segmentation), UterineAssist (segmentation algorithm), and deep learning-based view classification and structure detection algorithms.
source quote (p.7)
“The proposed HERA Z20 Diagnostic Ultrasound System; HERA Z20e Diagnostic Ultrasound System; HERA Z20s Diagnostic Ultrasound System; R20 Diagnostic Ultrasound System; HERA Z30 Diagnostic Ultrasound System; R30 Diagnostic Ultrasound System have included new AI-ML based software features: AbdomenAssist, BladderAssist, QualityCheck, PelvicAssist. [...] A deep learning-based view classification and structure detection algorithm was validated using 43,737 fetal ultrasound images collected at hospitals.”
Adaptive (vs locked)FDA source did not state this
PCCPFDA source did not state this
Cybersecurity addressedFDA source did not state this

Validation studies (7)

Retrospective clinical

n=62 patients · 2 site(s)

endpoints: Success Rate; Mean difference (Bias); 95% Limits of Agreement (LoA)

Retrospective clinical

n=63 patients · 2 site(s)

endpoints: Success Rate; Mean difference (Bias); 95% Limits of Agreement (LoA)

Retrospective clinical

n=283 patients

endpoints: Sensitivity; Specificity; Positive Predictive Value (PPV); Negative Predictive Value (NPV)

Retrospective clinical

n=40 patients · 1 site(s)

endpoints: Acceptance rates; Intraclass Correlation Coefficient (ICC); mean difference

Retrospective clinical

n=200 patients · 2 site(s)

endpoints: error rate; bias (Mean Difference); 95% confidence interval; 95% Limits of Agreement (LoA)

Retrospective clinical

n=60 patients · 2 site(s)

endpoints: average dice-score; average error range; Bland-Altman analysis (95% LoA, Mean difference)

Retrospective clinical

n=46 patients · 2 site(s)

endpoints: localization accuracy success rate

Reported performance (21 observations)

sensitivity0.81CI 0.81 to 1.00
source quote (p.14)
“The Sensitivity ranged from 0.81 to 1.00”
specificity0.85CI 0.85 to 1.00
source quote (p.14)
“the Specificity ranged from 0.85 to 1.00”
ppvas written: “QualityCheck - View Classification (Manual) Positive Predictive Value (PPV)”stated without valueCI 0.80 to 1.00
source quote (p.14)
“The Positive Predictive Value (PPV) ranged from 0.80 to 1.00”
npvas written: “QualityCheck - View Classification (Manual) Negative Predictive Value (NPV)”stated without valueCI 0.90 to 1.00
source quote (p.14)
“the Negative Predictive Value (NPV) ranged from 0.90 to 1.00.”
ppvas written: “QualityCheck - Structure Detection (Manual) Positive Predictive Value (PPV)”stated without valueCI 0.89 to 1.00
source quote (p.14)
“The PPV ranged from 0.89 to 1.00”
npvas written: “QualityCheck - Structure Detection (Manual) Negative Predictive Value (NPV)”stated without valueCI 0.99 to 1.00
source quote (p.14)
“the NPV ranged from 0.99 to 1.00.”
ppvas written: “QualityCheck - View Classification (LVA) Positive Predictive Value (PPV)”stated without valueCI 0.80 to 1.00
source quote (p.14)
“The PPV ranged from 0.80 to 1.00”
npvas written: “QualityCheck - View Classification (LVA) Negative Predictive Value (NPV)”stated without valueCI 0.91 to 1.00
source quote (p.14)
“the NPV ranged from 0.91 to 1.00.”
ppvas written: “QualityCheck - Structure Detection (LVA) Positive Predictive Value (PPV)”stated without valueCI 0.89 to 1.00
source quote (p.14)
“The PPV ranged from 0.89 to 1.00”
npvas written: “QualityCheck - Structure Detection (LVA) Negative Predictive Value (NPV)”stated without valueCI 0.99 to 1.00
source quote (p.14)
“the NPV ranged from 0.99 to 1.00.”
agreement_kappaas written: “PelvicAssist - LH measurement Intraclass Correlation Coefficient (ICC) for LH Area”0.9802
source quote (p.15)
“The resulting Intraclass Correlation Coefficient (ICC) for six measurements were LH Area (0.9802), LH Circ. (0.9837), LH AP (0.9910), LH Lat. (0.9536), Right LUG (0.9423), and Left LUG (0.9596).”
agreement_kappaas written: “PelvicAssist - LH measurement Intraclass Correlation Coefficient (ICC) for LH Circ.”0.9837
source quote (p.15)
“The resulting Intraclass Correlation Coefficient (ICC) for six measurements were LH Area (0.9802), LH Circ. (0.9837), LH AP (0.9910), LH Lat. (0.9536), Right LUG (0.9423), and Left LUG (0.9596).”
agreement_kappaas written: “PelvicAssist - LH measurement Intraclass Correlation Coefficient (ICC) for LH AP”0.991
source quote (p.15)
“The resulting Intraclass Correlation Coefficient (ICC) for six measurements were LH Area (0.9802), LH Circ. (0.9837), LH AP (0.9910), LH Lat. (0.9536), Right LUG (0.9423), and Left LUG (0.9596).”
agreement_kappaas written: “PelvicAssist - LH measurement Intraclass Correlation Coefficient (ICC) for LH Lat.”0.9536
source quote (p.15)
“The resulting Intraclass Correlation Coefficient (ICC) for six measurements were LH Area (0.9802), LH Circ. (0.9837), LH AP (0.9910), LH Lat. (0.9536), Right LUG (0.9423), and Left LUG (0.9596).”
agreement_kappaas written: “PelvicAssist - LH measurement Intraclass Correlation Coefficient (ICC) for Right LUG”0.9423
source quote (p.15)
“The resulting Intraclass Correlation Coefficient (ICC) for six measurements were LH Area (0.9802), LH Circ. (0.9837), LH AP (0.9910), LH Lat. (0.9536), Right LUG (0.9423), and Left LUG (0.9596).”
agreement_kappaas written: “PelvicAssist - LH measurement Intraclass Correlation Coefficient (ICC) for Left LUG”0.9596
source quote (p.15)
“The resulting Intraclass Correlation Coefficient (ICC) for six measurements were LH Area (0.9802), LH Circ. (0.9837), LH AP (0.9910), LH Lat. (0.9536), Right LUG (0.9423), and Left LUG (0.9596).”
diceas written: “UterineAssist - Segmentation test average dice-score of uterus of sagittal view”96.7
source quote (p.18)
“The average dice-score of uterus of sagittal view is 96.7%.”
diceas written: “UterineAssist - Segmentation test average dice-score of uterus of transverse view”95.8
source quote (p.18)
“The average dice-score of uterus of transverse view is 95.8%.”
diceas written: “UterineAssist - Segmentation test average dice-score of endometrium”86.8
source quote (p.18)
“The average dice-score of endometrium is 86.8%.”
f1as written: “UterineAssist - Size measurement test average error range of Measurements”stated without valueCI 0.87 – 1.79 mm
source quote (p.18)
“The average error range of Measurements: 0.87 – 1.79 mm.”
accuracyas written: “NerveTrack - Detection test localization accuracy success rate”92.19CI [90.03%, 94.34%]
source quote (p.19)
“The feature demonstrated a localization accuracy success rate of 92.19% (95% CI: [90.03%, 94.34%])”

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

85
recalls in product code, 24mo
554
MAUDE reports in code, 12mo
+85%
vs code's own 3-yr baseline
4
drift signals on this device
  • recall_reason_pattern

    Software/algorithm-related recall in product code IYN (Philips Ultrasound, LLC, initiated 2025-10-31): "Ultrasound system compatibility issues with Apple devices running iOS 18 may cause a failure to perform live imagining." Recalling firm is another firm in the same product code.

    first seen 2026-07-08 · recall res_event_number:97843

  • recall_reason_pattern

    Software/algorithm-related recall in product code IYN (GE Medical Systems, LLC, initiated 2025-09-18): "The Ultrasound-Guided Attenuation Parameter (UGAP) measurement data may display inaccurate values representing liver steatosis. This could potentially lead to inappropriate clinica" Recalling firm is another firm in the same product code.

    first seen 2026-07-08 · recall res_event_number:97726

  • recall_reason_pattern

    Software/algorithm-related recall in product code IYN (GE Medical Systems China Co., Ltd. Dev. Zone National Hi-Tech; No., initiated 2025-05-16): "GE HealthCare has become aware that the Estimated Fetal Weight (EFW) measurement data feature on the Versana Premier R3 and LOGIQ F R3 series ultrasound systems can display previou" Recalling firm is another firm in the same product code.

    first seen 2026-07-08 · recall res_event_number:96992

  • recall_reason_pattern

    Software/algorithm-related recall in product code IYN (Siemens Medical Solutions USA, Inc., initiated 2024-08-15): "If ultrasound systems with software, are changed from factory default to : 1) Milliliters per second (ml/sec, mL/sec) or 2) Milliliters per minute (ml/min, mL/min); then systems wi" Recalling firm is another firm in the same product code.

    first seen 2026-07-08 · recall res_event_number:95254

Recall and MAUDE counts are product-code-level (reports aren't reliably attributable to one device); a recall is shown as device-attributed only when the recall record itself lists this clearance number. 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).

Applicable FDA guidance — what the submission is measured against

FDA guidance documents and guiding principles applicable to 510(k) AI/ML devices in the Radiology panel. A curated reference index, not legal or regulatory advice — each item states its own status, and a draft is never binding.

Applicability is derived from the device's FDA advisory panel and pathway — cross-cutting guidances apply to every AI/ML device; panel-specific ones are flagged. Titles, dates, and links verified against fda.gov as of July 2026.

Constat Precedent · public FDA/CMS data · descriptive decision-support, not regulatory or reimbursement advice. Share this page: constat.dev/precedent/device/K252018