Search Results
Found 5 results
510(k) Data Aggregation
(86 days)
Vivid Pioneer is a general-purpose ultrasound system, specialized for use in cardiac imaging. It is intended for use by, or under the direction of a qualified and trained physician or sonographer for ultrasound imaging, measurement, display and analysis of the human body and fluid.
Vivid Pioneer is intended for use in a hospital environment including echo lab, other hospital settings, operating room, Cath lab and EP lab or in private medical offices. The systems support the following clinical applications:
Fetal/Obstetrics, Abdominal (including renal, GYN), Pediatric, Small Organ (breast, testes, thyroid), Neonatal Cephalic, Adult Cephalic, Cardiac (adult and pediatric), Peripheral Vascular, Musculo-skeletal Conventional, Musculo-skeletal Superficial, Urology (including prostate), Transesophageal, Transvaginal, Transrectal, Intra-cardiac, Intra-luminal and Interventional Guidance (including Biopsy, Vascular Access), Thoracic/Pleural and Intraoperative (vascular).
Modes of operation include: 3D, Real time (RT) 3D Mode (4D), B, M, PW Doppler, CW Doppler, Color Doppler, Color M Doppler, Power Doppler, Harmonic Imaging, Coded Pulse and Combined modes: B/M, B/Color M, B/PWD or CWD, B/Color/PWD or CWD, B/Power/PWD.
The proposed Vivid Pioneer is a general purpose, Track 3, diagnostic ultrasound system, which is primarily intended for cardiac imaging and analysis but also includes vascular and general radiology applications. It provides digital acquisition, processing, display and analysis capabilities. It consists of a mobile console with a height-adjustable control panel, color LCD touch panel, and a display monitor.
Vivid Pioneer includes a variety of electronic array transducers operating in linear, curved, sector/phased array, matrix array or dual array format, including dedicated CW transducers and real time 3D transducer. The proposed Vivid Pioneer can be used with the stated compatible OEM ICE transducers. The system includes capability to output data to other devices like printing devices.
The user-interface includes an operator control panel, a 23.8" High-Definition Ultrasound LCD type of display monitor (mounted on an arm for rotation and / or adjustment of height), a layout of pre-defined user controls (hard-keys) and a 15.6-inch multi-touch LCD panel with mode-and operation dependent soft-keys.
The operator panel also includes two loudspeakers for audio, shelves for convenient placement of papers or accessories, and 6 holders with cable management for the connected transducers.
The lower console is mounted on 4 rotational wheels with brakes, for ergonomic transport and safe parking. The lower console also includes all electronics for transmit and receive of ultrasound data, ultrasound signal processing, software computing, hardware for image storage, hard copy printing, and network access to the facility through both LAN and wireless (supported by use of a wireless LAN USB-adapter) connection.
This document describes the acceptance criteria and study proving the device meets the criteria for two AI features of the Vivid Pioneer Ultrasound System: AI Cardiac Auto Doppler and AI FlexiViews LAA.
1. Table of Acceptance Criteria and Reported Device Performance
AI Cardiac Auto Doppler
| Acceptance Criteria | Reported Device Performance |
|---|---|
| Feasibility score of > 95% | All clinical parameters, as performed by AI Cardiac Auto Doppler without user edits, passed the check for mean percent absolute error across all cardiac cycles against a threshold. This implies the accuracy threshold was met, which indirectly suggests successful feasibility to achieve this accuracy. |
| Expected accuracy threshold calculated as the mean absolute difference in percentage for each measured parameter. | All clinical parameters, as performed by AI Cardiac Auto Doppler without user edits, passed this check. |
| Mean percent absolute error across all cardiac cycles against a threshold. | All clinical parameters, as performed by AI Cardiac Auto Doppler without user edits, passed this check. |
| Consistent model performance across BMI groups (<25 and $\ge$ 25) with predefined metric quantifying agreement between manual and AI-derived peak velocities. | Tissue Doppler: Mean performance metric = -0.002 (SD = 0.077) for BMI < 25; -0.006 (SD = 0.081) for BMI $\ge$ 25.Flow Doppler: Mean performance metric = 0.021 (SD = 0.073) for BMI < 25; 0.003 (SD = 0.057) for BMI $\ge$ 25. |
AI FlexiViews LAA
| Acceptance Criteria | Reported Device Performance |
|---|---|
| Greater than 80% success rate of LAA region localization and landmark extraction | The model achieved a verification success rate of 85%, with a sensitivity of 84.91% and a specificity of 91.82%. Consistent model performance observed across TEE angles (0 to 100 degrees) with a success rate of 80% or higher. Strong model performance for individuals with a BMI above 25 (over 85% accuracy). |
2. Sample Size Used for the Test Set and Data Provenance
AI Cardiac Auto Doppler:
- Tissue Doppler test set: 4106 recordings from 805 individuals.
- Doppler Trace test set: 3390 recordings from 1369 individuals.
- Data Provenance: Retrospective, collected from USA (several locations), Australia, France, Spain, Norway, Italy, Germany, Thailand, Philippines.
AI FlexiViews LAA:
- Test set: 342 recordings from 84 individuals.
- Data Provenance: Retrospective, collected from USA, Norway, Italy, France, Philippines.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications
AI Cardiac Auto Doppler:
- Experts for annotations: Two cardiologists.
- Review panel for consensus: Five clinical experts.
- Qualifications: The document specifies "cardiologists" and "clinical experts" but does not explicitly state years of experience or board certification details.
AI FlexiViews LAA:
- Experts for annotations: Two cardiologists.
- Supervision for annotations: Two US certified clinicians.
- Review panel for consensus: Three clinical experts.
- Qualifications: The document specifies "cardiologists" and "US certified clinicians" and "clinical experts" but does not explicitly state years of experience or board certification details.
4. Adjudication Method for the Test Set
AI Cardiac Auto Doppler:
- Annotations were performed by two cardiologists.
- A review panel of five clinical experts provided feedback.
- Annotations were corrected (as needed) until a consensus agreement was achieved between the annotators and reviewers. This suggests an adjudication method aimed at reaching a single agreed-upon ground truth.
AI FlexiViews LAA:
- Annotations were performed by two cardiologists, supervised by two US certified clinicians.
- A review panel of three clinical experts provided feedback.
- Annotations were corrected (as needed) until a consensus agreement was achieved between the annotators and reviewers. Similar to Auto Doppler, this indicates a consensus-based adjudication.
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study
The provided text does not mention a multi-reader multi-case (MRMC) comparative effectiveness study to assess how much human readers improve with AI vs. without AI assistance for either AI Cardiac Auto Doppler or AI FlexiViews LAA. The evaluation focused on the standalone performance of the AI algorithms against expert-derived ground truth.
6. Standalone Performance (Algorithm Only)
Yes, standalone (algorithm only without human-in-the-loop performance) studies were done for both AI features.
- AI Cardiac Auto Doppler: Performance was evaluated based on the AI algorithm's measurements directly compared to expert-derived ground truth. The verification explicitly states "AI Cardiac Auto Doppler without user edits passed this check."
- AI FlexiViews LAA: The "model achieved a verification success rate of 85%" based on its localization and landmark extraction, directly reflecting standalone performance.
7. Type of Ground Truth Used
Expert Consensus.
For both AI Cardiac Auto Doppler and AI FlexiViews LAA, the ground truth was established through:
- Manual measurements/annotations performed by cardiologists.
- Assessment of Doppler/ECG signal quality.
- Supervision by US certified clinicians (for LAA).
- Review and consensus agreement among a panel of clinical experts.
8. Sample Size for the Training Set
AI Cardiac Auto Doppler:
- Tissue Doppler development dataset: 1482 recordings from 4 unique clinical sites.
- Doppler Trace development dataset: 2070 recordings from 4 unique clinical sites.
AI FlexiViews LAA:
- Total development dataset: 612 recordings from 5 unique clinical sites.
9. How the Ground Truth for the Training Set Was Established
The ground truth for the development (training/validation) datasets was established in the same manner as the ground truth for the test sets:
- For both AI Cardiac Auto Doppler and AI FlexiViews LAA:
- Annotators (cardiologists, supervised by US certified clinicians for LAA) performed manual measurements/annotations after assessing image quality (Doppler signal quality and ECG signal quality for Auto Doppler, LAA contour and specific points for FlexiViews LAA).
- Annotations followed US ASE (American Society of Echocardiography) based annotation guidelines.
- A review panel of clinical experts (five for Auto Doppler, three for FlexiViews LAA) provided feedback.
- Annotations were corrected (as needed) until a consensus agreement was achieved between the annotators and reviewers.
Ask a specific question about this device
(121 days)
The LOGIQ Totus is a general-purpose diagnostic ultrasound system intended for use by qualified and trained healthcare professionals for ultrasound imaging, measurement, display and analysis of the human body and fluid. LOGIQ Totus clinical applications include: Fetal / Obstetrics; Abdominal (including Renal, Gynecology/Pelvic); Pediatric; Small Organ (Breast, Testes, Thyroid); Neonatal Cephalic; Adult Cephalic; Cardiac(Adult and Pediatric); Peripheral Vascular; Musculo-skeletal Conventional and Superficial; Urology (including Prostate); Transrectal; Transvaginal.
Modes of operation include: B, M, PW Doppler, CW Doppler, Color Doppler, Color M Doppler, Power Doppler, Harmonic Imaging, Coded Pulse, 3D/4D Imaging mode, Elastography, Shear Wave Elastography, Attenuation Imaging, Contrast Enhanced Imaging and Combined modes: B/M, B/Color, B/PWD, B/Color/PWD, B/Power/PWD.
The system is intended to be used in Hospital or Clinical environments such as Intensive Care Unit(ICU, CVICU, CCU), Neonatal Intensive Care Unit(NICU), Pediatric Intensive Care Unit(PICU), Emergency Room, Operating Room, Outpatient Surgery Clinic, Radiology, Medical Office (Nurse Practitioner), Observational Units, Cath Lab, Clinic, Physician's Office, Labor/Deliver Unit and Oncology.
The LOGIQ Totus is full featured, Track 3 device, primarily intended for general purpose diagnostic ultrasound system which consists of a mobile console approximately 490mm wide(monitor width: 545mm), 835mm deep and 1415~1815mm high that provides digital acquisition, processing and display capability. The user interface includes a computer keyboard, specialized controls. 14-inch LCD touch screen and color 23.8-inch LCD & HDU image display.
The provided document contains information on two AI features: "Auto preset selection" and "Auto Abdominal Color Assistant". The acceptance criteria and study details for each are presented below.
Auto Preset Selection
1. A table of acceptance criteria and the reported device performance:
| Acceptance Criteria | Reported Device Performance (Summary Test Statistics) |
|---|---|
| The overall model success rate of the Abdomen, Air, Breast, Carotid, Leg, MSK, Scrotal, Thyroid and Carotid/Thyroid (Mixed) view suggestion is expected to be 80% or higher. | The document states, "The overall model success rate of the Abdomen, Air, Breast, Carotid, Leg, MSK, Scrotal, Thyroid and Carotid/Thyroid(Mixed) view suggestion is expected to be 80% or higher." This indicates the device met this criterion. |
2. Sample size used for the test set and the data provenance:
- Test Set Sample Size:
- Number of individual patients' images collected from: 50+ patients
- Number of samples (images): 330+ images
- Data Provenance:
- Country of Origin: USA (57%) and Australia (43%)
- Retrospective/Prospective: Not explicitly stated, but the mention of "data collection protocol was standardized" suggests it might be retrospective collection of pre-existing data, or a controlled prospective study.
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
- Number of Experts: Not explicitly stated how many individual sonographers/clinicians were involved.
- Qualifications of Experts: "Certified sonographer/clinician." No years of experience or specific board certifications are mentioned.
4. Adjudication method for the test set:
- Adjudication Method: "For the testing process, the results are generated by the AI software and the same are verified as Pass or Fail by a certified sonographer/clinician." This implies a single expert verification without explicit multi-reader adjudication (e.g., 2+1 or 3+1).
5. If a multi-reader multi-case (MRMC) comparative effectiveness study was done:
- No MRMC comparative effectiveness study was explicitly mentioned for this AI feature. The testing method described focuses on the AI's success rate verified by experts, not on comparing human performance with and without AI assistance.
6. If a standalone (i.e., algorithm only without human-in-the-loop performance) was done:
- Yes, the "overall model success rate" refers to the algorithm's performance in suggesting views. The "verified as Pass or Fail by a certified sonographer/clinician" is likely a post-hoc verification of the algorithm's output, rather than an interactive human-in-the-loop study.
7. The type of ground truth used:
- Type of Ground Truth: Expert verification by certified sonographer/clinician, determining if the AI's view suggestion was a "Pass or Fail". This can be considered a form of "expert consensus" or "expert truth" (albeit with a single expert verification as mentioned in point 4).
8. The sample size for the training set:
- Not explicitly stated. The document only mentions that "The exams used for test/training validation purpose are separated from the ones used during training process and there is no overlap between the two." The sample size (330+ images from 50+ patients) is specified for the test set.
9. How the ground truth for the training set was established:
- Not explicitly stated. The document mentions that the test data was independent of the training data but doesn't detail the ground truth establishment for the training set. It's generally assumed that ground truth for training data in such AI applications would also be established by experts.
Auto Abdominal Color Assistant
1. A table of acceptance criteria and the reported device performance:
| Acceptance Criteria | Reported Device Performance (Summary Test Statistics) |
|---|---|
| The overall model success rate of the Aorta, Kidney, Liver, GB and Pancreas view suggestion is expected to be 80% or higher. | The document states, "The overall model success rate of the Aorta, Kidney, Liver, GB and Pancreas view suggestion is expected to be 80% or higher." This indicates the device met this criterion. |
2. Sample size used for the test set and the data provenance:
- Test Set Sample Size:
- Number of individual patients' images collected from: 40 patients
- Number of samples (images): 280+ images
- Data Provenance:
- Country of Origin: USA (35%) and Australia (65%)
- Retrospective/Prospective: Similar to "Auto preset selection", not explicitly stated, but a "standardized data collection protocol" is mentioned.
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
- Number of Experts: Not explicitly stated how many individual sonographers/clinicians were involved.
- Qualifications of Experts: "Certified sonographer / clinician." No years of experience or specific board certifications are mentioned.
4. Adjudication method for the test set:
- Adjudication Method: "For the testing process, the results are generated by the AI software and the same are verified a Pass or Fail by a certified sonographer / clinician." This implies a single expert verification without explicit multi-reader adjudication.
5. If a multi-reader multi-case (MRMC) comparative effectiveness study was done:
- No MRMC comparative effectiveness study was explicitly mentioned for this AI feature.
6. If a standalone (i.e., algorithm only without human-in-the-loop performance) was done:
- Yes, the "overall model success rate" refers to the algorithm's performance. The "verified as Pass or Fail by a certified sonographer / clinician" is likely a post-hoc verification of the algorithm's output.
7. The type of ground truth used:
- Type of Ground Truth: Expert verification by certified sonographer/clinician, determining if the AI's view suggestion was a "Pass or Fail".
8. The sample size for the training set:
- Not explicitly stated. The document only mentions that "The exams used for test/training validation purpose are separated from the ones used during training process and there is no overlap between the two." The sample size (280+ images from 40 patients) is specified for the test set.
9. How the ground truth for the training set was established:
- Not explicitly stated. Similar to the "Auto preset selection" feature, it's assumed expert input would have been used for training data ground truth, but no details are provided.
Ask a specific question about this device
(127 days)
LOGIQ E10 is intended for use by a qualified physician for ultrasound evaluation of Fetal / Obstetrics; Abdominal (including Renal, Gynecology/Pelvic); Pediatric; Small Organ (Breast, Testes, Thyroid); Neonatal Cephalic; Adult Cephalic; Cardiac (Adult and Pediatric); Peripheral Vascular; Musculo-skeletal Conventional and Superficial; Urology (including Prostate); Transrectal; Transvaginal; Transesophageal and Intraoperative (Abdominal and Vascular).
Modes of operation include: B, M, PW Doppler, CW Doppler, Color Doppler, Color M Doppler, Power Doppler, Harmonic Imaging, Coded Pulse, 3D/4D Imaging mode, Elastography, Shear Wave Elastography, Attenuation Imaging and Combined modes: B/M, B/Color, B/PWD, B/Color/PWD, B/Power/PWD. The LOGIQ E10 is intended to be used in a hospital or medical clinic.
The LOGIQ E10 is a full featured, track 3, general purpose diagnostic ultrasound system which consists of a mobile console approximately 585 mm wide (keyboard), 991 mm deep and 1300 mm high that provides digital acquisition, processing and display capability. The user interface includes a computer keyboard, specialized controls, 12-inch highresolution color touch screen and 23.8-inch High Contrast LED LCD monitor.
Here's a breakdown of the acceptance criteria and study details for the "Auto Renal Measure Assistant" and "Auto Abdominal Color Assistant" features of the GE LOGIQ E10, based on the provided FDA 510(k) summary:
The document provides information for two distinct AI features: "Auto Renal Measure Assistant" and "Auto Abdominal Color Assistant". I will detail the information for each separately.
Auto Renal Measure Assistant
1. Table of Acceptance Criteria and Reported Device Performance
| Acceptance Criteria | Reported Device Performance |
|---|---|
| Longitudinal Model: | |
| Accuracy > 80% | 96.45% accuracy with 95% CI of ±1.26% |
| Transverse Model (Width Measurements): | |
| Accuracy > 70% | 92.94% accuracy with 95% CI of ±3.02% |
| Transverse Model (Height Measurements): | |
| Accuracy > 70% | 93.13% accuracy with 95% CI of ±3.63% |
2. Sample Size Used for the Test Set and Data Provenance
- Sample Size: 30 patients, resulting in 60 images (30 Longitudinal views and 30 Transverse views).
- Data Provenance:
- Country of Origin: USA (58%) and Japan (42%).
- Retrospective/Prospective: Prospectively collected.
3. Number of Experts Used to Establish Ground Truth for the Test Set and Qualifications
- Number of Experts: 2 Readers (certified sonographer/Clinician) and 1 Board Certified Nephrologist.
- Qualifications: "Certified sonographer/Clinician" for the initial readers; "Board Certified Nephrologist" for the arbitrator. Specific years of experience are not mentioned.
4. Adjudication Method for the Test Set
- Adjudication Method: A Board Certified Nephrologist arbitrated the ground truth between the two initial readers to establish the reference standard. This resembles a "2+1" or "tie-breaker" adjudication.
5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done
- No, a "Multi-Reader Multi-Case (MRMC) comparative effectiveness study" comparing human readers with and without AI assistance was not specified for this feature. The study focused on the algorithm's performance against expert-established ground truth.
6. If a Standalone (Algorithm Only) Performance Study Was Done
- Yes, the reported accuracies (96.45%, 92.94%, 93.13%) represent the standalone performance of the algorithm in measuring renal dimensions against the established ground truth.
7. The Type of Ground Truth Used
- Type of Ground Truth: Expert consensus (between two sonographers/clinicians, arbitrated by a nephrologist).
8. The Sample Size for the Training Set
- The document states that the verification data was acquired independently during validation after the development of the model. The sample size for the training set is not provided in this summary.
9. How the Ground Truth for the Training Set Was Established
- The method for establishing ground truth for the training set is not specified in this summary. Only the method for the independent verification (test) set is described.
Auto Abdominal Color Assistant / Auto Preset Assistant
(Note: The document lists "Auto Abdominal Color Assistant" and then immediately below it, and seemingly as a continuation or related feature, "Auto Preset Assistant" with similar testing information. I will treat them as two related or broadly similar features based on the provided structure.)
1. Table of Acceptance Criteria and Reported Device Performance
| Acceptance Criteria | Reported Device Performance |
|---|---|
| Auto Abdominal Color Assistant: | |
| Overall model success rate of Aorta, Kidney, Liver, GB, and Pancreas view suggestion is expected to be 80% or higher. | Not explicitly stated in a single number, but implied to meet criteria given context of a 510(k) summary. |
| Auto Preset Assistant: | |
| Overall model success rate of Abdomen, Air, Breast, Carotid, Leg, MSK, Scrotal, Thyroid and Carotid/Thyroid (Mixed) view suggestion is expected to be 80% or higher. | Not explicitly stated in a single number, but implied to meet criteria given context of a 510(k) summary. |
2. Sample Size Used for the Test Set and Data Provenance
- Auto Abdominal Color Assistant:
- Sample Size: 50+ patients, resulting in 1100+ images.
- Data Provenance:
- Country of Origin: USA (77%) and Australia (23%).
- Retrospective/Prospective: Not explicitly stated, but "collected from:" implies existing data or a mix.
- Auto Preset Assistant:
- Sample Size: 110+ patients, resulting in 2600+ images.
- Data Provenance:
- Country of Origin: USA (41.2%), Austria (3.8%), Australia (1.1%), Japan (41.3%), Italy (0.7%), and Greece (12%).
- Retrospective/Prospective: Not explicitly stated, but "collected from:" implies existing data or a mix.
3. Number of Experts Used to Establish Ground Truth for the Test Set and Qualifications
- Number of Experts: Unspecified number of "Readers (certified sonographer/Clinician)".
- Qualifications: "Certified sonographer/Clinician". Specific years of experience or precise number of experts not mentioned.
4. Adjudication Method for the Test Set
- Adjudication Method: Not specified. The summary states, "Readers (certified sonographer/Clinician) to ground truth the "anatomy" visible in static B-Mode image." It doesn't mention multiple readers for the same image or an adjudication process if there were discrepancies.
5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done
- No, an MRMC comparative effectiveness study was not specified for these features. The testing described is for the standalone algorithm's accuracy in view suggestion.
6. If a Standalone (Algorithm Only) Performance Study Was Done
- Yes. The core of the testing involved running the AI and comparing its predictions to the ground truth to calculate the accuracy of the algorithm against each class. This represents standalone performance.
7. The Type of Ground Truth Used
- Type of Ground Truth: Expert annotation by "certified sonographer/Clinician" on B-mode images ("anatomy visible").
8. The Sample Size for the Training Set
- The document states that exams used for test/training validation were separated with no overlap. However, the specific sample size for the training set is not provided.
9. How the Ground Truth for the Training Set Was Established
- The method for establishing ground truth for the training set is not specified. Only the method for the independent test set is described.
Ask a specific question about this device
(125 days)
The LOGIQ E10s and LOGIQ Fortis are intended for use by a qualified physician for ultrasound evaluation.
Specific clinical applications and exam types include: Fetal / Obstetrics; Abdominal (including Renal, Gynecology/Pelvic); Pediatric; Small Organ (Breast, Testes, Thyroid); Neonatal Cephalic; Adult Cephalic; Cardiac (Adult and Pediatric); Peripheral Vascular; Musculo-skeletal Conventional and Superficial; Urology (including Prostate); Transrectal; Transvaginal; Transesophageal and Intraoperative (Abdominal, Vascular).
Modes of operation include: B, M, PW Doppler, CW Doppler, Color Doppler, Color M Doppler, Power Doppler, Harmonic Imaging, Coded Pulse, 3D/4D Imaging mode, Elastography, Shear Wave Elastography, Attenuation Imaging and Combined modes: B/M, B/Color, B/PWD, B/Color/PWD, B/Power/PWD.
The LOGIQ E10s and LOGIQ Fortis are intended to be used in a hospital or medical clinic.
The LOGIQ E10s is a full featured, Track 3, general purpose diagnostic ultrasound system which consists of a mobile console approximately 585 mm wide (keyboard), 991 mm deep and 1300 mm high that provides digital acquisition, processing and display capability. The user interface includes a computer keyboard, specialized controls, 12-inch high-resolution color touch screen and 23.8-inch High Contrast LED LCD monitor.
The LOGIQ Fortis is a full featured, Track 3, general purpose diagnostic ultrasound system which consists of a mobile console approximately 575 mm wide (keyboard). 925 mm deep and 1300 mm high that provides digital acquisition, processing and display capability. The user interface includes a digital keyboard (physical keyboard as an option), specialized controls, 12inch high-resolution color touch screen and 23.8-inch High Contrast LED LCD monitor (or 23.8inch High Resolution LED LCD monitor as an option).
The provided text describes three AI features of the LOGIQ E10s and LOGIQ Fortis systems: Auto Renal Measure Assistant, Auto Abdominal Color Assistant, and Auto Preset Assistant. The information provided for each feature allows for a detailed breakdown of their acceptance criteria and the studies conducted to prove they meet these criteria.
Here's the requested information structured for clarity:
1. Table of Acceptance Criteria and Reported Device Performance
| AI Feature | Acceptance Criteria | Reported Device Performance |
|---|---|---|
| Auto Renal Measure Assistant | Longitudinal model accuracy for length measurements expected to be > 80%. Transverse model accuracy for width measurements expected to be > 70%. | Longitudinal model for length measurements: Average accuracy of 96.45% (95% CI: ±1.26%), average absolute error of 0.35cm (95% CI: ±0.12 cm). Transverse model for width measurements (first mention): Average accuracy of 92.94% (95% CI: ±3.02%), average absolute error of 0.38cm (95% CI: ±0.14 cm). Transverse model for width measurements (second mention, likely a typo/repetition): Average accuracy of 93.13% (95% CI: ±3.63%), average absolute error of 0.37cm (95% CI: ±0.14 cm). |
| Auto Abdominal Color Assistant | Overall model success rate for Aorta, Kidney, Liver, GB, and Pancreas view suggestion expected to be 80% or higher. | Specific accuracy percentages for each view are not individually reported in the summary, but the success rate is implied to have met or exceeded the 80% threshold, as the device is deemed substantially equivalent. The summary states "Calculated the accuracies of the algorithm against each class," which suggests these were evaluated. |
| Auto Preset Assistant | Overall model success rate for Abdomen, Air, Breast, Carotid, Leg, MSK, Scrotal, Thyroid, and Carotid/Thyroid (Mixed) view suggestion expected to be 80% or higher. | Specific accuracy percentages for each view are not individually reported in the summary, but the success rate is implied to have met or exceeded the 80% threshold, as the device is deemed substantially equivalent. The summary states "Calculated the accuracies of the algorithm against each class," which suggests these were evaluated. |
2. Sample Sizes and Data Provenance for Test Sets
- Auto Renal Measure Assistant:
- Test Set Sample Size: 30 patients, resulting in 60 samples (30 longitudinal views, 30 transverse views).
- Data Provenance: Prospective collection. Data from USA (58%) and Japan (42%).
- Auto Abdominal Color Assistant:
- Test Set Sample Size: 50+ patients, resulting in 1100+ images.
- Data Provenance: Not explicitly stated as retrospective or prospective, but collected from USA (77%) and Australia (23%).
- Auto Preset Assistant:
- Test Set Sample Size: 110+ patients, resulting in 2600+ images.
- Data Provenance: Not explicitly stated as retrospective or prospective, but collected from USA (41.2%), Austria (3.8%), Australia (1.1%), Japan (41.3%), Italy (0.7%), and Greece (12%).
3. Number of Experts and Qualifications for Ground Truth
- Auto Renal Measure Assistant:
- Number of Experts: 2 "Readers" and 1 "Board Certified Nephrologist" for arbitration.
- Qualifications: "certified sonographer/Clinician" for the two readers. "Board Certified Nephrologist" for the arbitrator.
- Auto Abdominal Color Assistant:
- Number of Experts: Unspecified number of "Readers".
- Qualifications: "certified sonographer/Clinician" for the readers.
- Auto Preset Assistant:
- Number of Experts: Unspecified number of "Readers".
- Qualifications: "certified sonographer/Clinician" for the readers.
4. Adjudication Method for Test Sets
- Auto Renal Measure Assistant:
- Method: A "Board Certified Nephrologist arbitrated the ground truth between the above two readers to establish the reference standard". This implies a 2+1 (two readers, one arbitrator) method.
- Auto Abdominal Color Assistant & Auto Preset Assistant:
- Method: The text states, "Readers (certified sonographer/Clinician) to ground truth the 'anatomy' visible in static B-Mode image." There is no mention of multiple readers or an arbitration process, implying no explicit inter-reader adjudication method was described beyond individual expert annotation.
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study
- No MRMC comparative effectiveness study was explicitly described in the provided text. The studies focus on the standalone performance of the AI algorithms against a derived ground truth, rather than comparing human reader performance with and without AI assistance. Therefore, no effect size for human readers' improvement with AI assistance is reported.
6. Standalone (Algorithm Only) Performance Study
- Yes, standalone (algorithm only) performance studies were done for all three AI features listed. The studies evaluate the accuracy or success rate of the AI algorithms in performing their intended functions (measurement, view suggestion) against an established ground truth, without a human-in-the-loop component being explicitly tested or reported.
7. Type of Ground Truth Used
- Auto Renal Measure Assistant: Expert Consensus, as it involved two readers and an arbitrator to establish the reference standard for measurements.
- Auto Abdominal Color Assistant & Auto Preset Assistant: Expert Annotation/Consensus, established by "Readers (certified sonographer/Clinician) to ground truth the 'anatomy'visible in static B-Mode image." While not explicitly stated as consensus among multiple readers, it is established by qualified experts.
8. Sample Size for Training Sets
- The training set sample sizes are not explicitly provided in the summaries for any of the AI features. The document only mentions that the "verification data was acquired independently during validation process after the development of the model," and "The exams used for test/training validation purpose are separated from the ones used during training process." This implies training data existed but its size is not detailed.
9. How Ground Truth for Training Sets Was Established
- The document does not explicitly describe how the ground truth for the training sets was established. It focuses primarily on the process for the test/validation sets. However, it can be inferred that a similar process involving expert clinicians/sonographers would have been used to establish ground truth for training data, as is common practice in medical imaging AI development.
Ask a specific question about this device
(171 days)
EchoPAC Software Only / EchoPAC Plug-in is intended for diagnostic review and analysis of ultrasound images, patient record management and reporting, for use by, or on the order of a licensed physician. EchoPAC Software Only / EchoPAC Plug-in allows post-processing of raw data images from GE ultrasound scanners and DICOM ultrasound images.
Ultrasound images are acquired via B (2D), M, Color M modes, Color, Power, Pulsed and CW Doppler modes, Coded Pulse, Harmonic, 3D, and Real time (RT) 3D Mode (4D).
Clinical applications include: Fetal/Obstetrics; Abdominal (including renal and GYN), Urology (including prostate), Pediatric; Small organs (breast, testes, thyroid), Neonatal and Adult Cephalic, Cardiac (adult and pediatric), Peripheral Vascular, Transesophageal (TEE), Musculo-skeletal Conventional, Musculo-skeletal Superficial, Transrectal (TR), Transvaginal (TV), Intraoperative (vascular), Intra-cardiac, and Intra-luminal.
EchoPAC Software Only / EchoPAC Plug-in provides image processing, annotation, analysis, measurement, report generation, communication, storage and retrieval functionality to ultrasound images that are acquired via the Vivid family of ultrasound scanners by GE Healthcare. The EchoPAC Software Only / EchoPAC Plug-in software is an integral component of each Vivid system, providing the post-acquisition image management and reporting functions of the scanner. EchoPAC Software Only will be offered as SW-only to be installed directly on customer PC hardware, and EchoPAC Plug-in will be offered as an accessory to image management workstations. EchoPAC Software Only / EchoPAC Plug-in is DICOM compliant, transferring images and data via LAN between scanners, hard copy devices, file servers and other workstations.
The provided text is a 510(k) Premarket Notification Submission for GE Healthcare's "EchoPAC Software Only / EchoPAC Plug-in." It describes the device, its intended use, and compares it to predicate devices. However, it explicitly states:
"The subject of this premarket submission, EchoPAC Software Only / EchoPAC Plug-in, did not require clinical studies to support substantial equivalence."
This means the document does not contain information about acceptance criteria, a study proving device performance against those criteria, sample sizes, expert involvement, or adjudication methods because clinical studies were not deemed necessary for this submission. The submission relies on non-clinical tests and comparison to previously cleared predicate devices.
Therefore, I cannot provide the requested information from the given text.
Ask a specific question about this device
Page 1 of 1