(86 days)
AFFINITY is a software application used to process, display, and manage nuclear medicine and other medical imaging data transferred from other workstations, PACS or acquisition stations. The information acquired from viewing the images is used, in conjunction with other patient related data, for diagnosis and monitoring of disease.
Warning! This application is not intended to replace visual assessment of tumors. The application is to provide pre-selection of lesions for visual confirmation and to provide consistency and reproducibility when assessing tumor response to treatment
Affinity is a software application used to process, display, analyse and manage nuclear medicine and other medical imaging data transferred from other workstations, PACS or acquisition stations. The information acquired from viewing the images is used, in conjunction with other patient related data, for diagnosis and monitoring of disease.
This application is not intended to replace visual assessment of tumors. The primary purpose of the application is to provide pre-selection of lesions for visual confirmation and to provide consistency and reproducibility when assessing tumor response to treatment.
Affinity can process data as whole body or constrained field of view (e.g. abdomen, brain) where scanning was performed with any of the following modalities PET/CT, MR or tomographic reconstructed SPECT/CT. The studies are read in DICOM format and if there are studies with different modalities, on the same patient, a co-registration and alignment is performed.
Software output - After the studies have been processed, these are presented visually in 2D or 3D for the user who can use tools to quantify the significant parameters SUV, SUVR, SUVbsa, SUVIbm, SUVbw, SUV Peak, SUV Mean, TLG and MTV.
Here's a breakdown of the acceptance criteria and study information for the Affinity device, based on the provided document:
Acceptance Criteria and Device Performance
| Acceptance Criteria | Reported Device Performance |
|---|---|
| Split tool intra/inter-variability (quantitative values for tumor assessment) | Variability was 0, except for in the 14-17th decimal place (indicating high reproducibility). |
| Hottest Connected SUV peak value accuracy (comparison to manual calculation) | Identical results across all 13 tested scenarios. |
Study Details
1. Sample Sizes and Data Provenance
- Test Set (for Split tool intra/inter-variability): 13 subjects (7 males, 6 females) who received F18FDG-PET/CT studies.
- Data Provenance: Not explicitly stated (e.g., country of origin). The study states "13 subjects... that received a F18FDG-PET/CT study," implying prospective data collection for this specific validation, but this is not explicitly confirmed.
- Test Set (for Hottest Connected SUV peak value accuracy): A single synthetic software-generated study containing 8 voxels.
- Data Provenance: Synthetic (software-generated).
2. Number of Experts and their Qualifications for Ground Truth
- For Split tool intra/inter-variability: No experts were used to establish ground truth in the traditional sense. The validation focused on the reproducibility of the tool itself when used by validators.
- For Hottest Connected SUV peak value accuracy: The "ground truth" was established by manual calculation. No specific number or qualifications of human experts were mentioned for this manual calculation in the provided text.
3. Adjudication Method for the Test Set
- For Split tool intra/inter-variability: Two validators repeated the process 5 times each. The variability between these repeat measurements and between the two validators was assessed. This is not a classic adjudication method for ground truth, but rather a reproducibility assessment.
- For Hottest Connected SUV peak value accuracy: Direct comparison to manual calculation. No adjudication method involving multiple human readers was described.
4. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study
- No, an MRMC comparative effectiveness study involving human readers improving with AI vs. without AI assistance was not conducted or reported in the provided text. The studies focused on the performance characteristics of specific software features.
5. Standalone Performance Study
- Split tool intra/inter-variability: Yes, the study assesses the standalone reproducibility of the "threshold/split tool" by having two validators use the tool on patient data.
- Hottest Connected SUV peak value accuracy: Yes, the study assesses the standalone accuracy of the "Hottest Connected SUV peak value" by comparing its output to manual calculations.
6. Type of Ground Truth Used
- For Split tool intra/inter-variability: The "ground truth" for this test was the assumption that the tool should produce consistent quantitative values. The study measured the reproducibility of these measurements rather than comparing them to a separate, external ground truth like pathology. The process involved defining regions based on a predefined threshold, splitting them, and then deleting specific anatomical regions (brain, heart, urinary bladder) and ensuring pathological areas connected to non-pathological ones were deleted based on clinical importance. This implies an implicit clinical understanding, but not a formally established "ground truth" for each specific lesion.
- For Hottest Connected SUV peak value accuracy: Manual calculation on a synthetic dataset. This serves as the "truth" against which the algorithm's output is compared.
7. Sample Size for the Training Set
- The document does not specify a sample size for the training set. The descriptions provided relate to verification and validation testing, not model training.
8. How the Ground Truth for the Training Set Was Established
- Since no information on a training set was provided, the method for establishing its ground truth is also not available in the document.
§ 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.