(127 days)
Quantra™ is a software application intended for use with images acquired using digital breast X-ray systems. Quantra calculates volumetric breast density as a ratio of fibroglandular tissue and total breast volume estimates; and area breast density as a ratio of fibroglandular tissue area and total breast area estimates. It segregates breast density into BI-RADS-like breast composition categories, which may be useful in the reporting of consistent breast composition values as mandated by certain state regulations. Quantra provides these numerical values for each image, breast, and subject, to aid radiologists in the assessment of breast tissue composition. Quantra produces adjunctive information; it is not an interpretive or diagnostic aid. Quantra runs on a Windows platform.
Quantra is a software application that estimates breast tissue volume and area density. The estimations are made from images acquired using digital breastray systems.
Here's a summary of the acceptance criteria and study details for the Hologic Quantra™ device, based on the provided text:
1. Table of Acceptance Criteria and Reported Device Performance
The provided text describes validation methods and comparisons rather than explicit, stand-alone acceptance criteria with quantitative thresholds. However, it implicitly demonstrates that the device's performance met the expectation of correlating with existing methods and providing similar distributions across diverse equipment.
| Metric/Measure | Acceptance Criteria (Implicit) | Reported Device Performance |
|---|---|---|
| Volumetric Breast Density (Vbd) | Correlation with predicate device (K082483, Quantra Volumetric Assessment) and MRI cases; comparison to mode BI-RADS density rating from 15 radiologists. | Demonstrated correlation with predicate device and MRI cases. |
| Area Breast Density (Abd) | Correlation with an expert's hand-drawn dense areas using predicate device (Sectra IDS5 Workstation); comparison to mode BI-RADS density rating from 15 radiologists. | Demonstrated correlation with expert hand-drawn dense areas on Sectra IDS5 Workstation. |
| BI-RADS-like Breast Composition (q_abd) | Comparison of continuous q_abd score to the mean BI-RADS values assigned by 15 radiologists; comparison across different FFDMs to ensure similar distributions. | Demonstrated consistency with mean BI-RADS values from 15 radiologists. |
| Consistency of Density Measures | Statistical evaluation of density measures between CC and MLO views of the same breast, and left/right breasts of the same women. | All Quantra density measures were evaluated statistically and found to be consistent. |
| Distribution Similarity | Comparison of Vbd, Abd, and q_abd values across large populations of cases from Hologic (Selenia, Dimensions), GE (Senographe, Essential), and Siemens (Mammomat Novation) FFDMs. | Distributions were similar across different FFDMs. |
2. Sample Size Used for the Test Set and Data Provenance
- Sample Size for Density Measures:
- Vbd/Abd/q_abd compared to BI-RADS: "a large database of cases" (exact number not specified).
- Vbd compared to MRI: "MRI cases of the same patients" (exact number not specified).
- Vbd/Abd/q_abd compared across FFDMs: "a large population of cases" from Hologic, GE, and Siemens (exact number not specified).
- Consistency (CC/MLO, Left/Right): "substantially large number of images" from Hologic, GE, and Siemens (exact number not specified).
- Data Provenance: Not explicitly stated (e.g., country of origin). The data included cases from Hologic (Selenia and Dimensions), GE (Senographe and Essential), and Siemens (Mammomat Novation) FFDMs, implying diversity in equipment. The data was "retrospective" as it was used for validation of an already developed software.
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications
- Number of Experts: 15 radiologists.
- Qualifications: "radiologists" - no further specific qualifications (e.g., years of experience, subspecialty) are provided in the document.
4. Adjudication Method for the Test Set
- The text describes comparing Quantra values to the "mode (most common) BI-RADS density rating from 15 radiologists" and the "mean value of the 15 radiologists." This indicates a form of consensus-based ground truth derived from multiple readers, but it doesn't specify a formal adjudication method like "2+1" or "3+1" (where differing opinions are resolved by a tie-breaker). It appears to use statistical aggregations (mode or mean) of expert ratings.
5. If a Multi Reader Multi Case (MRMC) Comparative Effectiveness Study was done
- No, a formal MRMC comparative effectiveness study "with AI vs without AI assistance" to measure the effect size of human readers improving with AI assistance was not explicitly described for this 510(k) submission. The studies detailed focus on validating the standalone performance of the Quantra software against existing clinical practices and predicate devices.
6. If a Standalone (i.e. algorithm only without human-in-the-loop performance) was done
- Yes, the primary focus of the performance/bench testing described is on the standalone performance of the Quantra algorithm. It generates volumetric and area breast density estimates and segregates breast density into BI-RADS-like categories without human input during the calculation process. Its output is "adjunctive information" for radiologists, implying it functions independently before a radiologist's assessment.
7. The Type of Ground Truth Used
- Expert Consensus: The "mode" and "mean" BI-RADS density ratings from 15 radiologists served as a form of expert consensus ground truth for comparison.
- Predicate Device Output: Results from the predicate device (Quantra Volumetric Assessment for Vbd, Sectra IDS5 Workstation with expert hand-drawn annotations for Abd) were used as ground truth for correlation.
- Other Imaging Modalities: MRI cases were used as ground truth for Vbd validation.
8. The Sample Size for the Training Set
- The sample size for the training set is not specified. The document mentions a "reference database" of "approximately 1,000 patients" used to create Vbd-score and Vfg-score (reflecting standard deviations from the mean in this database), but it's unclear if this was the training set or a separate reference population.
9. How the Ground Truth for the Training Set was Established
- The document does not explicitly describe how the ground truth for any potential training set was established. The focus of the "Performance/Bench Testing" section is on validation against established references and expert opinions, not on the training process itself. If the "reference database" of 1,000 patients was part of a training set, the method for establishing its ground truth is not detailed in the provided information.
§ 892.2050 Medical image management and processing system.
(a)
Identification. A medical image management and processing system is a device that provides one or more capabilities relating to the review and digital processing of medical images for the purposes of interpretation by a trained practitioner of disease detection, diagnosis, or patient management. The software components may provide advanced or complex image processing functions for image manipulation, enhancement, or quantification that are intended for use in the interpretation and analysis of medical images. Advanced image manipulation functions may include image segmentation, multimodality image registration, or 3D visualization. Complex quantitative functions may include semi-automated measurements or time-series measurements.(b)
Classification. Class II (special controls; voluntary standards—Digital Imaging and Communications in Medicine (DICOM) Std., Joint Photographic Experts Group (JPEG) Std., Society of Motion Picture and Television Engineers (SMPTE) Test Pattern).