Search Results

The Multi-SPECT is intended for use as an accessory to an dual headed nuclear medicine gamma camera (for example - the ADAC Vertex Dual Headed Nuclear Medicine Camera). It is to be used to perform static, dynamic and/or gated Single Photon Emission Computed Tomography clinical nuclear medicine images of the heart. It is indicated for use to produce images showing the location and distribution of radioisotopes in the human body for interpretation by a physician.

The product consists of two multi-pinhole collimators attached to the two heads of a dual headed gamma camera (eg ADAC (now Philips) Vertex Gamma Camera). The collimators are designed to perform myocardial perfusion studies. Each collimator consists of a 2 x 3 array of pinholes. Each pinhole is focused on the patient's heart, such that the radioisotope distribution in the heart is seen from 18 different view points. This multiplicity of views allows for tomographic reconstruction of the radioisotope distribution in the heart. Acquisition is done in list-mode, retaining position, energy and patient physiological data. The raw listmode data is framed and corrected for scatter, then reconstructed into a three dimensional matrix. It can then be analyzed by previously cleared software packages (eg Emory Cardiac Toolbox™ - K040141).

Here's a detailed breakdown of the acceptance criteria and the study proving the device meets them, based on the provided 510(k) summary:

1. Table of Acceptance Criteria and Reported Device Performance:

The document primarily focuses on demonstrating substantial equivalence to predicate devices through technical specifications and NEMA test results, rather than explicitly stating pre-defined "acceptance criteria" in a go/no-go fashion. However, the comparisons imply that the Multi-SPECT's performance should be at least comparable to and ideally better than the predicate devices, particularly in sensitivity and resolution.

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

• Quantitative Performance: The "test set" for the quantitative performance metrics (resolution and sensitivity) consisted of NEMA NU 1 phantom(s) or equivalent, following NEMA guidelines. This is a standardized phantom, not a patient-specific dataset.
• Clinical Images: "A set of three clinical images was also examined." The document does not specify the origin (country, retrospective/prospective) of these three clinical images. It's likely these were illustrative cases rather than a statistically powered clinical trial.
• Data Provenance: Phantom data is laboratory-generated. The provenance of the three clinical images is not detailed.

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

• Quantitative Performance: For the NEMA phantom tests, the "ground truth" is inherent in the phantom's design and known properties, measured directly by the testing equipment. No human experts are explicitly mentioned for establishing this ground truth.
• Clinical Images: The document mentions "A set of three clinical images was also examined." However, it does not state how the ground truth for these images was established, nor how many experts were involved or their qualifications. Given the small number, these were likely for visual inspection rather than a formal validation study with expert ground truth.

4. Adjudication Method for the Test Set:

• Quantitative Performance: No adjudication method is specified for the NEMA phantom tests, as these are objective measurements.
• Clinical Images: No adjudication method is described for the "set of three clinical images."

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

• No, an MRMC comparative effectiveness study was NOT done. The submission focuses on demonstrating substantial equivalence through technical specifications and phantom studies, and a very limited review of three clinical images. There is no mention of human readers evaluating images with and without AI assistance to quantify improvement.

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

• Yes, in essence, standalone performance was assessed. The NEMA phantom tests and direct measurements of resolution and sensitivity are "standalone" in the sense that they evaluate the physical imaging system (collimator + gamma camera + reconstruction software) without human interpretation as the primary endpoint. The "algorithm" here refers to the entire imaging and reconstruction pipeline.

7. The Type of Ground Truth Used:

• Quantitative Performance: The ground truth for the quantitative NEMA phantom tests is the known physical properties of the phantom and the direct measurements derived from it.
• Clinical Images: The type of ground truth for the "set of three clinical images" is not specified.

8. The Sample Size for the Training Set:

• Not Applicable / Not Provided. This device is a hardware accessory (collimator) and associated reconstruction software algorithms. The document does not describe the use of machine learning or deep learning models that would typically require a training set of images with established ground truth. The reconstruction software uses "underlying mathematical algorithms" which are likely model-based or analytical, not learned from a large dataset in the way modern AI is trained.

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

• Not Applicable / Not Provided. As per point 8, there's no indication of a training set or a machine learning approach in the sense that would require image data with established ground truth for model training. The algorithms are based on established mathematical principles for SPECT reconstruction that incorporate pinhole geometry.

Ask a specific question about this device