Search Results

The iSR'obot MRI-US Fusion is a software application to be used by Clinicians in the clinic or hospital for 2-D and 3-D visualization, image registration, and fusion of Magnetic Resonance and Ultrasound images for mapping planning information of the prostate gland and region of interest. The software features also include multi-modality data communication, surface and volume rendering, segmentation, multi-planar reconstruction, organ delineation, region of interest delineation, landmark selection, measurements and data reporting.

The iSR'obot MRI-US Fusion (UroFusion) is a software application intended for use by Clinicians or Radiologist for 2-D and 3-D visualization, image registration and fusion of Magnetic Resonance and Ultrasound images for mapping planning of the prostate gland and region of interest, such as lesions, to provide "MRI-3D model" imaqe information. The "MRI-3D model" image information produced by this software acts as inputs to the iSR'obot Mona Lisa which allows the import of this "MRI-3D model" image information; and fusion of this model information together with the live ultrasound 3D-model image information.

Leveraging on the information available from both Magnetic Resonance (MRI) & Ultrasound modalities concurrently, the fusion results enable the clinicians to visualize the prostate and the region of interest (lesions); thus enabling fewer and more accurate targeted prostate biopsies to be taken as compared with "blind" saturated biopsies with ultrasound guidance alone.

UroFusion software system includes the following features:

• Access and display medical imaging studies from MRI DICOM data. .
• Provide for 2D contouring /3D modelling of the prostate gland. .
• Provide for 2D contouring / 3D modelling of tumour / lesions within the . gland.
• Provide for saving of patient "MRI-3D model" information together with ● the patient's MRI DICOM data.
• Provide for importing of patient "MRI-3D model" information together with relevant MRI DICOM data.
• Provide for the fusion of "MRI-3D model" information with the live . ultrasound 3D- model information for subsequent operations to be executed in iSR'obot Mona Lisa.

UroFusion can be deployed and utilized in commercially available computer platforms and operating systems; or as a standalone system.

Here's an analysis of the acceptance criteria and the study that proves the device meets the acceptance criteria, based on the provided document:

1. Table of Acceptance Criteria and Reported Device Performance:

The document does not explicitly define a table of acceptance criteria with numerical targets. Instead, it relies on demonstrating substantial equivalence to a predicate device (Multi-Modality Image Fusion, K120187) and performing verification and validation activities, including non-clinical performance evaluation and clinical evaluation through literature review.

The "reported device performance" is primarily qualitative, focusing on whether the UroFusion successfully fulfills its intended functions and is comparable to the predicate. The key performance demonstrated in the non-clinical testing is the measurement of "Hausdorff distance," which is a metric for comparing two point sets or shapes, indicating how far two surfaces are from each other. However, a specific acceptance criterion for Hausdorff distance (e.g., "Hausdorff distance < X mm") is not provided.

Therefore, a direct table of acceptance criteria and reported performance with quantitative metrics cannot be fully populated from this document in the traditional sense of a performance study with pre-defined thresholds.

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

• Non-Clinical Testing:

  • Sample Size: "2 phantoms" were used.
  • Data Provenance: Not explicitly stated, but phantoms are simulated-use objects.

• Clinical Testing: The document does not describe a clinical study with a specific test set of patients or medical images in the traditional sense. Instead, it relies on a literature review.

  • Sample Size:
    • "6 relevant papers" were evaluated.
    • "3 of them were original clinical studies" (implies these studies had their own patient cohorts, but the size of those cohorts is not specified in this document).
    • The other 3 papers were review papers.
    • "2 statements were made by physicians who had experiences with our devices." (These are anecdotal endorsements, not a structured clinical test set).
  • Data Provenance: The literature search would likely encompass data from various countries based on the publications. It is implicitly retrospective as it reviews existing published data.

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

• Non-Clinical Testing (Phantoms):
  • Ground truth for phantoms is typically inherent in their design or established through precise physical measurements, not by expert consensus in the same way as biological data. The document does not specify experts for establishing ground truth for the phantoms.
• Clinical Testing (Literature Review):
  • For the reviewed clinical studies, the ground truth would have been established by the authors of those individual studies. The document does not provide details on the number or qualifications of experts involved in those studies.
  • For the 2 statements from physicians, their qualifications are "physicians who had experiences with our devices," but no further details are given about their experience level or the method by which their "statements" constituted ground truth.

4. Adjudication Method for the Test Set:

• Non-Clinical Testing (Phantoms): Not applicable for phantom-based technical performance testing as described.
• Clinical Testing (Literature Review): No formal adjudication method is mentioned. The "evaluation" of the papers was done by the manufacturer, Biobot Surgical Pte Ltd, as part of their submission strategy.

5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance:

No. The document explicitly states: "The clinical evaluation of UroFusion was carried out using the literature search route... no clinical trials are required." Therefore, no MRMC comparative effectiveness study was performed or described in this submission.

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

• The non-clinical testing, particularly the measurement of Hausdorff distance on phantoms, is generally a form of standalone technical performance testing of the algorithm's accuracy in tasks like image registration and segmentation. While humans would operate the software, the measurement of performance metrics like Hausdorff distance is an objective assessment of the algorithm's output against a known ground truth (the phantom's design).
• However, the document does not explicitly present a dedicated "standalone performance" study that would typically involve a specific dataset processed solely by the algorithm and evaluated against ground truth. The focus is on the device as a "software application to be used by Clinicians."

7. The Type of Ground Truth Used:

• Non-Clinical Testing: For the phantom studies, the ground truth would have been the known physical properties and dimensions of the phantoms.
• Clinical Testing (Literature Review): The ground truth in the reviewed clinical studies would vary depending on the study design, but for prostate lesion detection and biopsy guidance, it often includes pathology reports from biopsies as the gold standard, and potentially expert consensus on imaging in some cases. The document does not specify the ground truth types for the individual studies reviewed.

8. The Sample Size for the Training Set:

The document does not provide information about a training set or its sample size. This is typical for a 510(k) submission for certain types of software, especially those that primarily focus on image processing and fusion and are seeking substantial equivalence to a predicate, rather than developing novel diagnostic algorithms that require extensive training. The device's technological basis is described as "employing the same fundamental scientific technology as that of its predicate device," suggesting it might not involve a machine learning model that requires a distinct "training set."

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

Since no training set is described, the method for establishing its ground truth is not provided.

Ask a specific question about this device