Search Results

The cardiac position emission tomography (PET) analysis tool, HeartSee is a software package intended for use by nuclear medicine and nuclear cardiology physicians and technologists to facilitate image interpretation. Archiving of output data will be supported for clinical diagnostics, quality control, and research.

HeartSee is intended for processing of DICOM images, visual analysis and quantification of relative myocardial tracer uptake, and quantification of absolute myocardial blood flow and CFR when applied to diagnostic cardiac PET images in patients with suspected or known coronary artery disease.

The cardiac positron emission tomography (PET) analysis tool HeartSee is a software package intended for use by nuclear medicine and nuclear cardiology physicians and technologists to facilitate image interpretation. Archiving of output data will be supported for clinical diagnostics, quality control, and research.

HeartSee contains two fundamental components. First, the software can import cardiac PET images in DICOM format from any camera manufacturer. These images can be reoriented to cardiac axes to produce standard tomographic and topographic displays of relative uptake. A trained, licensed physician can interpret these processed images as per standard practice.

Second, the CFR software can quantify absolute myocardial blood flow per unit tissue (cc/min/gm) in stress and rest PET cardiac images and quantitatively assess the coronary flow reserve (CFR).

To compute coronary flow reserve (CFR) - the ratio of increased blood flow (stress) to baseline blood flow (rest) – three inputs are required: integrated arterial activity in the early part after bolus injection, average myocardial activity in the late part after bolus injection, and correction factors for partial volume effects of the PET scanner. The first number comes from a region of interest (ROI) drawn in the thoracic aorta or left atrium on images taken soon after radionuclide bolus administration. The second number comes from the topographic maps of myocardial uptake acquired later after radiotracer injection. The third number varies by PET camera and will be initialized in a user preference file.

HeartSee is a software package that uses standard, industrial computing hardware and applications.

This device is HeartSee, a cardiac PET analysis software.

Please note that the provided document is a 510(k) summary, which often provides less granular detail on study designs compared to full study reports or publications. Therefore, some information may not be explicitly available and will be noted as such.

Here's a breakdown of the requested information:

1. Table of Acceptance Criteria and Reported Device Performance

The 510(k) summary does not explicitly state formal acceptance criteria with numerical thresholds. Instead, it relies on demonstrating substantial equivalence to predicate devices (K090178 and K113754). This typically means the new device performs similarly well or not worse than the predicates.

However, the document mentions specific validation studies for its two main functionalities:

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

• Test Set Sample Size: 82 patients.
• Data Provenance: The document does not explicitly state the country of origin. It mentions data from patients with coronary artery disease and "cardiac PET images acquired at our institution." Given the contact information for K. Lance Gould, M.D. at the University of Texas Medical School at Houston, it is highly likely the data is from the USA.
• Retrospective or Prospective: The wording "MBF from 82 patients processed by the new algorithm was compared to MBF previously calculated by an independent algorithm" suggests that the data was retrospective, as the "previously calculated" values imply existing data.

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

The ground truth for the quantitative component (MBF and CFR) was established by an "independent algorithm" (presumably cfrQuant, predicate K113754), not directly by human experts in the sense of adjudication. Therefore, the concept of "number of experts" for establishing this specific ground truth isn't applicable in the traditional sense here.

For the visual interpretation component, the document states: "A trained, licensed physician can interpret these processed images as per standard practice." This implies standard clinical practice where a single (or multiple, depending on practice) trained, licensed physician would interpret the images. However, no specific number of experts or their qualifications for a formal ground truth establishment for this visual component are detailed in the 510(k) summary.

4. Adjudication Method for the Test Set

For the quantitative component (MBF and CFR), the adjudication method was a comparison against a "previously calculated" value from an "independent algorithm" (predicate K113754). This is a direct algorithmic comparison, not a human adjudication method like 2+1 or 3+1.

For the visual interpretation component, no specific adjudication method is described within the context of a validation study for the 510(k) submission.

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, an MRMC comparative effectiveness study involving human readers with and without AI assistance is not described in this 510(k) summary. The validation focused on the standalone performance of the algorithm and its correlation with a predicate algorithm.

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

Yes, a standalone performance study was done for the quantitative components (MBF and CFR). The study directly compared the HeartSee algorithm's output for MBF and CFR to the output of an "independent algorithm" (the predicate device, cfrQuant). The correlations (r=0.99) demonstrate the standalone performance of the HeartSee algorithm.

7. The Type of Ground Truth Used

• For MBF and CFR Quantification: The ground truth was "previously calculated" values from an "independent algorithm" (the predicate device, cfrQuant). This can be classified as a form of algorithmic reference standard or established method comparison.
• For Visual Interpretation/Image Reorientation: The implied ground truth relies on standard clinical practice and expert interpretation, but a formal ground truth definition for a validation study is not detailed.

8. The Sample Size for the Training Set

The 510(k) summary does not provide information on the sample size used for the training set for the HeartSee device. This type of detail is often not included in 510(k) summaries.

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

The 510(k) summary does not provide information on how the ground truth for the training set was established, as it does not describe the training process or dataset.

Ask a specific question about this device