Search Results

The Pristina Serena 3D option provides the three-dimensional location of target lesions, using information obtained from Digital Breast Tomosynthesis (DBT) images. This information provides guidance for a variety of minimally invasive or interventional procedures in the breast such as: vacuum assisted biopsy, core biopsy, pre-surgical localization (e.q. hookwire), and fine needle aspirations (FNA).

Pristina Serena 3D is DBT Biopsy System for Senographe Pristina. It is an additional option that builds upon the Pristina Serena device (GE Healthcare Stereotaxy biopsy option for Senographe Pristina platform). Pristina Serena was cleared on May 14, 2018 (K173576).

Pristina Serena 3D enables biopsy medical application to be done using Digital Breast Tomosynthesis (DBT) images.

The Pristina Serena 3D add-on includes the following items: Software: A new software version for the Senographe Pristina platform which includes software to manage the Pristina Serena 3D option. Labeling for the Biopsy DBT (3D) Medical application. Pristina Serena 3D option is compatible with previously installed Senographe Pristina systems. Pristina Serena 3D does not require any hardware modification on the Senographe Pristina platform. The hardware that was cleared on Pristina Serena 3D (K173576) was also not modified.

The Pristina Serena 3D device is designed to provide three-dimensional localization of target lesions using Digital Breast Tomosynthesis (DBT) images to guide minimally invasive breast procedures.

Here's an analysis of the acceptance criteria and the studies that prove the device meets these criteria:

1. Acceptance Criteria and Reported Device Performance

The provided document does not explicitly state a table of quantifiable acceptance criteria with corresponding performance metrics like sensitivity, specificity, or accuracy for lesion detection or biopsy targeting. Instead, the "Determination of Substantial Equivalence" section emphasizes the device's conformance to standards, successful verification/validation testing, and demonstrated performance according to specifications.

However, based on the technology description and comparison with reference devices, the key performance criteria for a biopsy guidance system generally revolve around accuracy of lesion localization and image quality.

• "Both 3D localization features calculate a three dimensional location for percutaneous placement for biopsy, pre-surgical localization or treatment devices." (referring to the predicate and reference devices, implying similar performance)
• "same stated accuracy" (when comparing with Hologic's Affirm) |
  | Image Quality for Biopsy Procedures | -"Since Pristina Serena 3D option uses the Image chain of the Senographe Pristina platform, the image quality of images during the biopsy procedure is equivalent to that of standard screening/diagnostic images of Senographe Pristina 3D."
• "Non-Clinical Data - Image Quality and Dose test that demonstrates that images acquired with Pristina Serena 3D are of same quality as images acquired with Senographe Pristina at similar dose levels." |
  | Dose Management | -"a single DBT acquisition allows for both confirming appropriate breast positioning and for lesion targeting (as opposed to a Stereotaxy procedure that requires three X-ray acquisitions: one scout image plus a stereotaxic pair). This translates into dose reduction to patient."
• "uses a single AOP mode available when acquiring biopsy images. This AOP mode provides the same parameters as the Senographe Pristina 3D AOP mode called "STD+" (P130020/S003 3D STD+ on Senographe Pristina 3D submission)." |
  | Mechanical Accuracy/Precision of Biopsy Device Holder | -"motorized and takes into account the geometry of the biopsy needles, so when the biopsy device holder is in place, the user introduces the needle in the breast until reaching the mechanical stop of the biopsy device holder..." |
  | Usability/Functionality | -"Pristina Serena 3D Design and Usability Validation Report which contains evidence of validation of claims and performance bench testing"
• "allows two different needle approaches for Biopsy: vertical and horizontal" |
  | Safety and Compliance | -"successfully completed required design control testing per GE Healthcare's quality management system."
  -"designed and will be manufactured under the Quality System Regulations of 21CFR 820 and ISO 13485." |

2. Sample Size for the Test Set and Data Provenance

The document does not specify a distinct "test set" in terms of clinical images or patient data with a defined sample size for evaluating the device's diagnostic or interventional accuracy. The evaluation appears to be based on non-clinical data, engineering bench performance testing, and design/usability validation.

• "Non-Clinical Data - Image Quality and Dose test": No sample size or data provenance (e.g., country of origin, retrospective/prospective) is provided for this test. It focuses on demonstrating equivalent image quality and dose to the predicate.
• "Pristina Serena 3D Design and Usability Validation Report": No sample size for user studies or specific details on data provenance are provided. This likely involved internal validation testing.

3. Number of Experts Used to Establish Ground Truth and Qualifications

Since specific clinical test sets are not detailed in the summary, information regarding the number of experts, their qualifications, or how they established ground truth for these tests is not provided. The assessment appears to rely on simulated use testing and technical performance specifications rather than clinical reads.

4. Adjudication Method for the Test Set

As there is no described clinical "test set" involving expert interpretation, an adjudication method (such as 2+1 or 3+1) is not applicable or described in this document.

5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study

A MRMC comparative effectiveness study that demonstrates how much human readers improve with AI vs. without AI assistance was not mentioned in the provided 510(k) summary. This device is an accessory for biopsy guidance and not an independent AI-powered diagnostic tool for image interpretation.

6. Standalone (Algorithm Only Without Human-in-the-Loop Performance)

The device acts as a guidance system. Its primary function is to accurately calculate 3D coordinates for a biopsy. The technology description focuses on the system's ability to determine 3D location and guide the biopsy needle, not an algorithm that performs standalone interpretation. Therefore, a standalone performance evaluation in the typical sense of an AI diagnostic algorithm is not discussed. The device's "performance" is inherently linked to its interaction with the imaging platform and the biopsy accessories.

7. Type of Ground Truth Used

Based on the documentation, the "ground truth" for evaluating the targeting accuracy in the non-clinical and bench testing would likely be phantom measurements or simulated lesion placements with known coordinates. For image quality, the ground truth would be established through technical image quality metrics and comparison to established standards for the predicate device. For usability, the ground truth would come from user feedback and task completion analysis. Clinical outcomes data are not mentioned as being used for these specific tests.

8. Sample Size for the Training Set

The document does not describe the specific underlying algorithms or models that would require a "training set" in the context of machine learning. The device is described as using principles of DBT image analysis to determine 3D locations. If any internal algorithms that contribute to coordinate calculations or image processing were developed using data, the sample size of such a training set is not disclosed.

9. How the Ground Truth for the Training Set Was Established

As no "training set" for a machine learning model is explicitly mentioned or described, the method for establishing its ground truth is not provided. The device functionality appears to be based on established physics and geometric principles of imaging and targeting rather than a data-driven learning approach from a large labeled dataset.

Ask a specific question about this device