Search Results

EMBOSPHERE® Microspheres are indicated for embolization of hypervascular tumors and arteriovenous malformations.

EMBOSPHERE® Microspheres are spherical microbeads for arterial embolization, made of acrylic polymer impregnated with gelatin. They are delivered with the help of a microcatheter in an amount appropriate to the area to be embolized. Six ranges of EMBOSPHERE® Microspheres are available in order to allow the physician to choose the calibration necessary for the vessel being embolized:

• 40-120 microns
• 100-300 microns
• 300-500 microns
• 500-700 microns
• 700-900 microns
• 900-1200 microns

The provided text describes a 510(k) summary for Biosphere Medical, Inc. EMBOSPHERE® Microspheres, an artificial embolization device. The document focuses on demonstrating substantial equivalence to predicate devices rather than providing a detailed study design with acceptance criteria and device performance metrics in the typical format of a clinical trial for a standalone AI device.

Therefore, many of the requested elements for describing acceptance criteria and a study that proves the device meets them, especially those pertaining to AI/ML device evaluation (like sample size for test sets, data provenance, number of experts, adjudication methods, MRMC studies, standalone performance, and training set details), are not applicable or cannot be extracted from this document.

However, I can extract information related to the performance testing employed to demonstrate substantial equivalence for the EMBOSPHERE® Microspheres.

1. Table of Acceptance Criteria and Reported Device Performance

The document does not present specific quantitative acceptance criteria or a direct comparison of device performance against such criteria in a table format. Instead, it lists types of performance testing that support substantial equivalence. The "performance" in this context refers to the device's characteristics and its comparable behavior to predicate devices and established clinical practice, rather than specific performance metrics (e.g., accuracy, sensitivity, specificity) typically associated with AI/ML device evaluations.

2. Sample size used for the test set and the data provenance

• Sample Size:
  • Animal Testing: The document does not specify the sample size (number of animals) used.
  • Clinical Data (European experience): The specific sample size for the European clinical data (referenced as "clinical data describing the European experience") is not provided in this summary.
  • Comparative Clinical Study: The specific sample size for the "clinical study comparing EMBOSPHERE® Microspheres to PVA for preoperative embolization of meningiomas" is not provided in this summary.
• Data Provenance:
  • Animal Testing: Implied to be laboratory-based (likely in a controlled research setting).
  • Clinical Data: "European experience." This is retrospective in the sense that it relies on past clinical use.
  • Comparative Clinical Study: The citation indicates a study by Bendszus M, Klein R, Burger R, et al.: "Efficacy of trisacryl gelatin microspheres versus polyvinyl alcohol particles in the prooperative embolization of meningiomas." This was published in AJNR in Feb 2000, suggesting a prospective clinical trial.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts

This information is not provided in the document. For device evaluations like this (medical products, not AI), the concept of "ground truth" and expert consensus as used in AI/ML performance testing is not directly applicable in the same way. Clinical outcomes and imaging results would be assessed by medical professionals, but their number and specific qualifications for establishing a "ground truth" for a test set are not detailed.

4. Adjudication method

This information is not provided and is not typically relevant for the type of device evaluation described in this 510(k) summary. Adjudication methods (like 2+1 or 3+1) are common in AI/ML studies where multiple human readers' interpretations need to be reconciled to establish a ground truth.

5. If a multi-reader multi-case (MRMC) comparative effectiveness study was done, and what was the effect size of how much human readers improve with AI vs without AI assistance

• MRMC Study: No, a multi-reader multi-case (MRMC) comparative effectiveness study involving human readers with and without AI assistance was not done. This is not an AI/ML device.
• The document mentions a "clinical study comparing EMBOSPHERE® Microspheres to PVA for preoperative embolization of meningiomas." This is a comparative effectiveness study between two embolization materials, not an MRMC study related to AI. Therefore, there is no effect size related to AI assistance.

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

This is not applicable as the EMBOSPHERE® Microspheres are a medical device (microspheres for embolization), not an algorithm or AI system. Their performance is inherently tied to their physical and biological interaction within the patient and their delivery by a physician.

7. The type of ground truth used

For the clinical and animal studies, the "ground truth" would implicitly be derived from:

• Clinical Outcomes Data: Observed patient outcomes (e.g., successful embolization, tumor reduction, AVM occlusion, complications, tissue reaction) as assessed by clinicians and confirmed through imaging or other diagnostic methods.
• Pathology: Potentially from tissue samples in animal studies to assess tissue reaction and depth of penetration.
• Imaging Data: Radiographic imaging results (e.g., visualization of embolization process, extent of embolization).

8. The sample size for the training set

This is not applicable as the EMBOSPHERE® Microspheres are a physical medical device, not an AI/ML model that requires a training set. The development of the device involved bench testing and iterative design, but not "training data" in the AI sense.

9. How the ground truth for the training set was established

This is not applicable for the reason stated above (not an AI/ML model requiring a training set).

Ask a specific question about this device