Search Results
Found 1 results
510(k) Data Aggregation
(140 days)
Insight BD
Insight BD is a software application for the volumetric density assessment of digital x-ray images of the breast to aid health care professionals in the assessment of breast tissue composition. Insight BD is not a diagnostic aid and should be used only as adjunctive information when the final assessment of breast density category is made visually by medical professionals.
The Siemens volumetric breast density analysis application software Insight BD provides volumetric assessment of digital x-ray images of the breast, including both digital mammograms and raw tomosynthesis projection images.
The assessment consists of generating and evaluating density maps where the value at each pixel represents the thickness of fibroglandular tissue above that pixel. From those density maps the software calculates quantitative values, namely
- . volume of fibroglandular tissue in the breast in cm³
- breast volume in cm³ .
- . volumetric breast density in %
- . breast density grade
The breast density grade is computed by thresholding of the volumetric breast density and conforms to BI-RADS® 5th Edition breast composition categories. The masking effect of dense fibroglandular tissue is automatically included in the results.
Insight BD is a software option for the Siemens digital mammography systems. Insight BD may be installed either on Siemens released PC hardware or on the Siemens digital mammography systems. Breast Density results are stored in the DICOM header and can be used for further documentation.
Here's an analysis of the acceptance criteria and the study proving the device meets those criteria, based on the provided text for the Siemens Insight BD (K172832) device.
Device Name: Insight BD
K Number: K172832
The provided text focuses on demonstrating substantial equivalence to a predicate device (Volpara Imaging Software, K152028). While it mentions various performance assessments, it does not explicitly state specific quantitative "acceptance criteria" and "reported device performance" in a structured table as requested. Instead, it describes various tests performed to demonstrate consistency, reproducibility, and equivalence to the predicate.
Therefore, the table below will be constructed based on the types of performance assessments described, with the "Acceptance Criteria" inferred from the need to show "similar" or "expected" results, and "Reported Device Performance" reflecting the general statement that the device was found compliant or similar to expectations/predicate.
1. Table of Acceptance Criteria and Reported Device Performance
| Acceptance Criteria (Inferred from study description) | Reported Device Performance (Summary from text) |
|---|---|
| Measurement Accuracy: Insight BD estimates are consistent with known values from breast phantoms. | "Measurement accuracy was assessed by comparing Insight BD estimates with known values from breast phantoms." (Implies satisfactory results) |
| Reproducibility (Left/Right, CC/MLO): Results for left and right breasts, and CC and MLO views, are similar. | "Reproducibility was assessed by running Insight BD over substantial data sets and the results for left and right breasts and CC and MLO views were compared to confirm that the results were similar for each view and each breast." |
| Reproducibility (DBT vs. FFDM): Results for the same woman imaged in tomosynthesis (DBT) and mammography (FFDM) modes are similar. | "Reproducibility was assessed by running Insight BD over substantial data sets where the same woman had been imaged on Siemens mammography systems in both tomosynthesis (DBT) and mammography (FFDM) modes, and the results were compared to confirm they were similar." |
| Reproducibility (Mammography vs. PRIME): Results for the same woman imaged in mammography and PRIME modes are similar. | "Reproducibility was assessed by running Insight BD over substantial datasets where the same women had been imaged on Siemens units running in both mammography and PRIME modes and the results were compared to confirm they were similar." |
| Consistency (with age): Results show an expected decrease in breast density with age. | "Consistency was assessed by running Insight BD over substantial datasets where the women's age was known and results were compared with the expected and known decrease in breast density with age." |
| Relationship to Visual Assessment: Insight BD output aligns with BI-RADS® 5th Edition density categories from qualified radiologists. | "Relationship to visual assessment was assessed by running Insight BD over x-ray images for which a BI-RADS® 5th Edition density category was available from MQSA qualified radiologists followed by a comparison of the two sets of data." (Implies a favorable comparison) |
| Equivalence to Predicate: Insight BD results are similar to the predicate device (Volpara). | "Equivalence was assessed by running both the predicate device Volpara and Insight BD over substantial datasets and the results were compared to confirm they were similar." |
2. Sample Size Used for the Test Set and Data Provenance
The document consistently uses the phrase "substantial data sets" to describe the sample sizes for reproducibility and equivalence testing, but does not provide specific numbers for the exact sample size of the test sets.
Data Provenance:
- Country of Origin: Not explicitly stated, but the manufacturing site is Forchheim, Germany (Siemens Healthcare GmbH). The lack of specific country for the data implies it could be multi-national or from Siemens' internal clinical collaborators.
- Retrospective or Prospective: Not explicitly stated. Given the use of "available from MQSA qualified radiologists" and "same woman had been imaged," it strongly suggests a retrospective collection of existing image data.
3. Number of Experts Used to Establish Ground Truth for the Test Set and Their Qualifications
The document mentions "MQSA qualified radiologists" for establishing BI-RADS® 5th Edition density categories based on visual assessment.
- Number of Experts: Not specified. The document states "a BI-RADS® 5th Edition density category was available from MQSA qualified radiologists," implying multiple radiologists might have contributed to the "available" categories, but it doesn't specify if this was a consensus reading or individual readings used for comparison.
- Qualifications: "MQSA qualified radiologists." This implies they are certified to interpret mammograms in the U.S. and have relevant clinical experience in breast imaging.
4. Adjudication Method for the Test Set
The document does not describe a specific adjudication method (e.g., 2+1, 3+1) for resolving disagreements among experts during the ground truth establishment for the visual assessment comparison. It simply states that BI-RADS categories were "available from MQSA qualified radiologists," which could imply a single read or an already adjudicated consensus (though not explicitly stated).
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study
No, a MRMC comparative effectiveness study was not explicitly done/described in the context of comparing human readers with AI assistance versus without AI assistance.
The study focuses on validating the standalone performance of Insight BD and its relationship to visual assessment and equivalence to a predicate. It does not evaluate the impact of the AI on human reader performance.
6. Standalone Performance Study (Algorithm Only)
Yes, standalone performance was done. The entire "Performance Data" Section 9 describes tests where the Insight BD software (algorithm only) was run on datasets to assess:
- Measurement accuracy (against phantoms)
- Reproducibility across different views and imaging modes
- Consistency with known biological trends (age-related density decrease)
- Relationship to visual assessment (comparison of algorithm output to radiologist-assigned BI-RADS categories)
- Equivalence to the predicate device
These are all evaluations of the algorithm's output directly.
7. Type of Ground Truth Used
Multiple types of "ground truth" or reference standards were used:
- Known values from breast phantoms: For measurement accuracy. This is an objective, engineered truth.
- Expected and known medical trends (decrease in breast density with age): For consistency analysis. This represents biological/clinical truth.
- BI-RADS® 5th Edition density categories available from MQSA qualified radiologists: For comparing the algorithm's output to human visual assessment. This is an expert consensus/visual assessment ground truth, though the consensus process isn't detailed.
- Predicate device output: For assessing equivalence. This uses another validated device's output as a reference, implying its output serves as a de facto "truth" for comparative purposes.
8. Sample Size for the Training Set
The document does not specify the sample size used for the training set for the Insight BD algorithm. It only discusses the "Performance Data" for the test (validation) phase.
9. How Ground Truth for the Training Set Was Established
The document does not describe how the ground truth for the training set was established. This information is typically proprietary to the developer and not always fully disclosed in 510(k) summaries, which tend to focus on the validation/testing to demonstrate substantial equivalence.
Ask a specific question about this device
Page 1 of 1