Search Results

The Siemens SOMATOM Definition AS Open systems are intended to produce cross-sectional images of the body by computer reconstruction of x-ray transmission data from either the same axial plane taken at different angles or spiral planes* taken at different angles.

(*spiral planes: the axial planes resulted from the continuous rotation of detectors and x-ray tube, and the simultaneous translation of the patient.)

New software version syngo® VA48 (SOMARIS/7 VA48) is a command-based program used for patient management, data management, X-ray scan control, image reconstruction, and image archive/evaluation that will be available on the SOMATOM Definition AS Open Computed Tomography systems. syngo® VA48 (SOMARIS/7 VA48) is a further development to the SOMARIS/7 operating software cleared as part of the predicate device.

syngo® VA48 is scanner platform software that supports the following device features:

• 1). New system scanner software version synqo® VA48 (SOMARIS/7 VA48) which includes:
  • -Respiratory Analysis of Respiratory Rate & Pitch Adjustment -FAST 3D Reconstruction (FAST 3D Align)
  • -Multiphase reconstruction with extended Field of View
  • -FAST DE Results (Dual Energy PACS-ready images)
  • -FAST contact
  • -Iterative Reconstruction with extended Field of View
  • -OEM Varian RGSC Online Mode
  • -Full 4D Lung Scan
  • -Applications at CT syngo.via client
  • -TrueD 4D Viewer
• 2). ADMIRE Iterative Reconstruction (Option)
• 3). iMAR Improved Metal Artifact Reduction (Option)

There are no modifications to the hardware of the device.

This document, a 510(k) Summary for the Siemens SOMATOM Definition AS Open CT system with software version syngo® VA48, primarily focuses on demonstrating substantial equivalence to a predicate device rather than presenting a detailed clinical study demonstrating improved human reader performance with AI assistance. It describes software updates and their verification and validation against technical standards.

Therefore, many of the requested details, such as specific acceptance criteria for AI performance, clinical study design for improved human reader performance, sample sizes for test sets in an MRMC study, expert qualifications for ground truth in a clinical context, or the effect size of AI assistance on human reader performance, are not explicitly available within this document. This submission is for a computed tomography x-ray system, and the "AI" or "machine learning" components mentioned (e.g., ADMIRE Iterative Reconstruction, iMAR Improved Metal Artifact Reduction) are features of the imaging system and reconstruction algorithms, not typically standalone AI interpretation tools that would undergo an MRMC study in the way a diagnostic AI would.

Given the information provided, here's what can be extracted and inferred, with limitations noted:

1. Table of Acceptance Criteria and Reported Device Performance:

The document doesn't provide a table of performance acceptance criteria in the sense of a diagnostic AI's clinical metrics (e.g., sensitivity, specificity, AUC). Instead, it refers to acceptance criteria for software specifications and conformance to technical standards.

2. Sample Size Used for the Test Set and Data Provenance:

The document refers to "Performance tests" and "Non-clinical tests (integration and functional)" but does not specify a sample size for a clinical test set of patient data, nor its provenance (country, retrospective/prospective). This type of information would be expected for a diagnostic AI device, not typically for a CT system software update focusing on features like iterative reconstruction or metal artifact reduction.

3. Number of Experts and Qualifications for Ground Truth:

Not applicable to this type of submission. There is no mention of human experts establishing ground truth for a diagnostic test set in the context of this 510(k). The "ground truth" here likely refers to technical specifications and expected performance characteristics of the CT system and its software, validated through engineering and phantom testing, rather than clinical interpretation.

4. Adjudication Method for the Test Set:

Not applicable. No clinical adjudication process is described as there isn't a stated clinical test set requiring human interpretation for ground truth.

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

No, an MRMC study was explicitly not done or described in this document. This regulatory submission is for a computed tomography x-ray system, specifically a software update (syngo® VA48) that includes features like iterative reconstruction (ADMIRE) and metal artifact reduction (iMAR). These are image processing and acquisition technologies, not AI-driven diagnostic assistance tools designed to change human reader performance in a comparative effectiveness study.

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

The document describes "Performance tests" to "test the functionality of the SOMATOM Definition AS Open configured with software version syngo VA48." While not using the term "standalone performance" in the context of a diagnostic AI, the testing described appears to be algorithm-only, focused on the technical performance and functional verification of the software features themselves (e.g., image reconstruction quality, artifact reduction effectiveness) against technical standards. "The results of these tests demonstrate that the subject device performs as intended."

7. Type of Ground Truth Used:

The ground truth for this submission appears to be based on:

• Technical specifications and engineering standards: Conformance to various IEC and NEMA standards (e.g., IEC 60601-2-44, IEC 61223-3-5, NEMA XR-25, DICOM).
• Internal software specifications and functional requirements: "The test results show that all the software specifications have met the acceptance criteria."
• Risk analysis and mitigation: "The Risk analysis was completed, and risk control implemented, to mitigate identified hazards." (The "ground truth" here is the identified hazards and their successful mitigation).

There is no mention of expert consensus, pathology, or outcomes data as a ground truth for clinical performance in this document.

8. Sample Size for the Training Set:

Not applicable. This document is not describing a machine learning model that was "trained" on a dataset in the way a diagnostic AI would be. The software updates described (like ADMIRE and iMAR) are based on algorithms and iterative processes, but the traditional concept of a "training set" for a deep learning model isn't presented here.

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

Not applicable, as there is no described training set for an AI model in this document.

Ask a specific question about this device

The Siemens SOMATOM Definition AS Open systems are intended to produce cross-sectional images of the body by computer reconstruction of x-ray transmission data from either the same axial plane taken at different angles or spiral planes* taken at different angles.

(*spiral planes: the axial planes resulted from the continuous rotation of detectors and x-ray tube, and the simultaneous translation of the patient.)

The Siemens SOMATOM Definition AS Open is a whole body X-ray Computed Tomography System. The SOMATOM Definition AS Open produces CT images in DICOM format, which can be used by post-processing applications commercially distributed by Siemens and other vendors.

The system software is a command-based program used for patient management, data management, X-ray scan control, image reconstruction, and image archive/evaluation. The new version of system software, syngo® CT 2013B (SOMARIS/7 VA46A), supports the following features:

• MARIS (Metal Artifact Reduction in Image Space) A image . reconstruction mode designed to reduce image artifacts caused by metal
• HD FoV Pro (HD FoV 2.0) Designed to enable a more reliable . visualization of the skin line of human body parts located outside of the standard field of view
• t-MIP -- Image manipulation method for arithmetic operations which allows . the calculation of temporal Maximum or Minimum Intensity Projection (MIP) images from a set of series.

Here's a breakdown of the acceptance criteria and the study information for the SOMATOM Definition AS Open configured with software version syngo® CT 2013B (SOMARIS/7 VA46A), based on the provided text:

Important Note: The provided document is a 510(k) summary for a medical device which is largely about demonstrating "substantial equivalence" to a predicate device. This type of submission often focuses on verifying that new features don't introduce new safety or effectiveness concerns, rather than conducting a full-scale clinical trial to prove a specific level of diagnostic performance against a robust ground truth. As such, some of the requested information (especially about specific performance metrics tied to acceptance criteria, MRMC studies, and detailed ground truth establishment for clinical effect) might not be explicitly present or as detailed as in a typical in vitro diagnostic (IVD) or AI-only software submission.

1. Table of Acceptance Criteria and Reported Device Performance:

2. Sample Size Used for the Test Set and Data Provenance:

• Sample Size (Clinical Tests): Not specified. The document states "Clinical tests were performed... to validate the performance of the MARIS algorithm" and "These tests include testing of the metal artifact reduction capabilities of MARIS in different clinical scenarios." However, the number of patients, scans, or images is not provided.
• Data Provenance: Not explicitly stated (e.g., country of origin). The tests were "clinical tests," implying real patient data. It is highly likely to be retrospective clinical data, as typical for 510(k) submissions focusing on software improvements, but this is not explicitly confirmed.

3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications:

• The document mentions "clinical tests" for MARIS validation, but it does not specify how ground truth was established for these tests, nor does it mention the number or qualifications of experts involved in any ground truth assessment. In the context of metal artifact reduction, "ground truth" might be subjective visual assessment by radiologists if not compared to a gold standard imaging modality.

4. Adjudication Method for the Test Set:

• The document does not specify any adjudication method (e.g., 2+1, 3+1). Given the lack of detail on expert involvement, it's unlikely a formal adjudication process was described for the submission.

5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study was Done:

• No, an MRMC comparative effectiveness study was not done (or at least not described in this summary). The studies mentioned focus on validating the performance of features (MARIS, HD FoV Pro) in the device itself, not on comparing human reader performance with and without AI assistance from this specific device.

6. If a Standalone (i.e. algorithm only without human-in-the-loop performance) was Done:

• The document describes "bench tests were performed to verify and validate the performance of the MARIS and HD FoV Pro (HD FoV 2.0) features," which are likely standalone algorithm evaluations using phantoms or controlled datasets.
• "Clinical tests" were also performed for the MARIS algorithm, which would involve the algorithm processing clinical data. While these involve physicians interpreting the output of the CT system, the focus of the "clinical tests" was on validating the algorithm's performance (e.g., artifact reduction), rather than a human reading study. So, in terms of the algorithm itself, yes, standalone performance was assessed.

7. The Type of Ground Truth Used:

• For the "bench tests" of MARIS and HD FoV Pro, the ground truth would likely be phantom-based measurements and technical specifications. Phantoms provide a known, controlled environment to assess image quality, artifact reduction, and field of view accuracy.
• For the "clinical tests" of MARIS, the document does not explicitly state the type of ground truth used. In the context of artifact reduction, it could involve visual assessment by clinicians comparing images with and without MARIS, or a comparison to an established 'gold standard' image if available (e.g., a non-metallic scan of the same area if feasible). However, no specifics are provided.

8. The Sample Size for the Training Set:

• The document does not specify the sample size for any training set. This is not uncommon for 510(k) submissions where the software updates are incremental and rely on established engineering practices, rather than a deep learning model requiring a distinct, large training dataset. The MARIS algorithm is described as an "image reconstruction mode," implying an algorithmic approach rather than a machine learning model that undergoes explicit "training" on a labeled dataset.

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

• As the document does not mention a training set in the context of machine learning, there is no information on how its ground truth was established. For algorithmic development, the "training" (design and tuning) is based on engineering principles, image science, and potentially smaller, internally-derived datasets with known properties or simulated artifacts.

Ask a specific question about this device