(59 days)
The equipment is intended to obtain position emission tomography (PET) images of parts of the fit in the patient aperture (e.g. head or breast) to detect abnormal pattern of distribution of a positron emitting radiopharmaceutical. This information can assist in diagnosis, therapeutic outcome assessment.
The SET-5002 PET scanner is designed to image the patient's breast or head which is placed in the aperture provided. The scanner can be easily switched from head to breast mode and back again with the push of a button on the control panel. When positrons are emitted from the radioactive drug administered to the patient, they annihilate electrons in body tissue. That process emits two annihilation photons (hereinafter "gamma rays") with an energy of 511 keV in the 180-degree direction. The PET System collects data by simultaneously counting both of these gamma rays using detectors arranged in a circular configuration. After collecting data for the specified scan time, an image is then reconstructed from that data based on the quantitative distribution of the radioactive drug. The following corrections are necessary for obtaining a quantitative distribution image: Normalization, Attenuation Correction, Scatter Correction, Random Correction, Decay Correction, Standard Time Correction, Count Losses Correction, and Cross Calibration. Operation of the system requires trained persons. The system is designed to be operated in a controlled medical environment such as a hospital or clinic. The unit is supplied with a computer which acquires and stores patient images. The scintillator type is Lutetium. The detectors are SiPM (Silicon photomultipliers).
The provided text describes the Shimadzu SET-5002 PET scanner and its premarket notification to the FDA. While it mentions performance testing and a clinical reader study, the document does not provide explicit acceptance criteria in a quantitative format, nor does it detail a comparative effectiveness study (MRMC) with human readers or a standalone AI performance study.
The information primarily focuses on demonstrating substantial equivalence to a predicate device (BBX-PET Scanner K210450) based on technological characteristics and general performance metrics, rather than specific acceptance criteria for a new AI/software component within the device.
However, based on the available information, I can infer and reconstruct some aspects relevant to performance validation as described:
Derived Acceptance Criteria and Reported Device Performance (Inferred):
Since no explicit quantitative acceptance criteria for image quality or diagnostic accuracy are listed, the "acceptance" is implied by the study's conclusion: "acceptable diagnostic results" and "supports the determination of substantial equivalence."
| Acceptance Criteria (Inferred from Study Conclusion) | Reported Device Performance |
|---|---|
| Overall image quality demonstrates acceptable diagnostic results for brain images. | Reader attested that her assessments of image quality of the SET-5002 demonstrated acceptable diagnostic results for brain images. |
| Overall image quality demonstrates acceptable diagnostic results for breast images. | Reader attested that her assessments of image quality of the SET-5002 demonstrated acceptable diagnostic results for breast images. |
| Performance of the new device is as safe, as effective, and performs as well as or better than the predicate device. | "The performance testing data supports the safety of the device and the validation testing demonstrate that new device should performs as intended in the specified use." Also, "From the result of our risk analysis, software verification and testing discussed above, it is our conclusion that the SET-5002 device is substantially equivalent to the legally marketed predicate devices. Therefore, new device is as safe, as effective, and performs as well as or better than the predicate device." |
Study Details based on the provided text:
-
Sample Size and Data Provenance:
- Test Set Sample Size: 18 images (12 brain images and 6 breast images).
- Data Provenance: U.S. clinical site. The document states, "A clinical reader study using the SET-5002 was conducted at U.S. clinical site." The retrospective or prospective nature is not explicitly mentioned but typically clinical reader studies for regulatory submission are on retrospectively collected and de-identified data.
-
Number of Experts and Qualifications:
- Number of Experts: One (singular "radiologist" and "reader attested").
- Qualifications: "Board certified radiologist." Specific years of experience are not mentioned.
-
Adjudication Method:
- Method: None specified in the document. The study was conducted with a single board-certified radiologist reviewing the images.
-
Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study:
- Was it done?: No. The study described is a single-reader study rather than a multi-reader, multi-case comparative effectiveness study.
- Effect Size: Not applicable, as no MRMC study was conducted comparing human readers with and without AI assistance. The study assessed "overall image quality... demonstrated acceptable diagnostic results" rather than an improvement metric for human readers.
-
Standalone (Algorithm Only) Performance Study:
- Was it done?: No. The described "clinical reader study" involved a human radiologist reviewing images produced by the SET-5002. There is no mention of an AI algorithm within the SET-5002 that would have standalone performance measured independent of human interpretation. The SET-5002 is a PET imaging system, not an AI diagnostic software.
-
Type of Ground Truth Used:
- Type: The ground truth for the reader study appears to be the assessment by the board-certified radiologist herself regarding "acceptable diagnostic results" and "PET imaging findings related to the progression of dementia, tumor detection, and the extent of spread." It's not stated that the images were confirmed against a definitive ground truth like pathology or long-term clinical outcomes. It implies the radiologist's assessment of image quality for diagnostic purposes served as the "ground truth" for the device's performance in this context.
-
Training Set Sample Size:
- Sample Size: Not applicable. The SET-5002 is described as an "Emission Computed Tomography System" (PET scanner), not an AI/machine learning software that requires a training set. The performance validation for such a device typically involves demonstrating image quality and physical performance metrics (resolution, sensitivity, etc.) and confirming suitability for diagnostic use by experts.
-
How Ground Truth for Training Set was Established:
- Method: Not applicable, as there is no mention of a training set for an AI/ML algorithm within the device.
§ 892.1200 Emission computed tomography system.
(a)
Identification. An emission computed tomography system is a device intended to detect the location and distribution of gamma ray- and positron-emitting radionuclides in the body and produce cross-sectional images through computer reconstruction of the data. This generic type of device may include signal analysis and display equipment, patient and equipment supports, radionuclide anatomical markers, component parts, and accessories.(b)
Classification. Class II.