Search Results
Found 1 results
510(k) Data Aggregation
(113 days)
This device is a digital radiography/fluoroscopy system used in a diagnostic and interventional angiography configuration. The system is indicated for use in diagnostic and angiographic procedures for blood vessels in the heart, brain, abdomen and lower extremities.
aEvolve Imaging is an imaging chain intended for adults, with Artificial Intelligence Denoising (AID) designed to reduce noise in real-time fluoroscopic images and signal enhancement algorithm, Multi Frequency Processing (MFP).
The Alphenix, INFX-8000V/B, INFX-8000V/S, V9.6 with αEvolve Imaging (FOV Extension), is an interventional X-ray system with a floor mounted C-arm as its main configuration. An optional ceiling mounted C-arm is available to provide a bi-plane configuration where required. Additional units include a patient table, X-ray high-voltage generator and a digital radiography system. The C-arms can be configured with designated X-ray detectors and supporting hardware (e.g. X-ray tube and diagnostic X-ray beam limiting device). With Alphenix, INFX-8000V/B, INFX-8000V/S, V9.6 with αEvolve Imaging (FOV Extension), the αEvolve Imaging feature now supports 12-inch, 10-inch, and 3-inch fields of view (FOV) for imaging in adult patients. The αEvolve imaging chain incorporates Artificial Intelligence Denoising (AID) for real-time fluoroscopic noise reduction, as well as Multi-Frequency Processing (MFP), a signal enhancement algorithm.
The provided 510(k) clearance letter describes performance testing for an interventional fluoroscopic X-ray system called "Alphenix, INFX-8000V/B, INFX-8000V/S, V9.6 with αEvolve Imaging (FOV Extension)". This device includes "Artificial Intelligence Denoising (AID)" and a "Multi Frequency Processing (MFP)" signal enhancement algorithm. The testing compares the subject device's αEvolve Imaging chain with AID to the predicate device's "super noise reduction filter (SNRF)".
Here's an analysis of the acceptance criteria and the study details:
1. Table of Acceptance Criteria and Reported Device Performance
The acceptance criteria are generally defined as the subject device performing "equivalent to or better than" the predicate, or "significantly better" (p < 0.05), or showing "no unexpected distortions" and "maintained or improved" performance.
| Performance Test | Acceptance Criteria | Reported Device Performance |
|---|---|---|
| Binning Mode Bench Test Results | ||
| Change in Image Level, Noise Magnitude and SNR | Image-level similarity using TOST, and noise magnitude and SNR properties equivalent to or better than the predicate (one-sided Student's t-test). | Noise and SNR properties of the subject device were equivalent to or better than those of the predicate. |
| Noise Power Spectrum (NPS) | Absence of unexpected distortions (e.g., spikes). | Both NPS curves were smooth and free of unexpected distortions. The subject IP chain exhibited a flatter NPS curve, with lower noise at spatial frequencies below ~0.6 cycles/mm and slightly higher noise above that range. |
| Noise Texture via Kurtosis | Subject IP chain's kurtosis being significantly closer to 3 than the predicate (p < 0.05) in most test cases. | The subject IP chain consistently met this criterion, indicating a more Gaussian-like noise distribution, while the predicate exhibited higher kurtosis. |
| Modulation Transfer Function (MTF) | MTF curve showed reduced over-enhancement and no unexpected distortions. | Both MTF curves were smooth and free of unexpected distortions. The subject IP chain applied more moderate enhancement compared to the predicate's higher MTF peak. |
| Noise Equivalent Quanta (NEQ) | Subject IP chain demonstrating higher NEQ in the low to mid spatial frequency range compared to the predicate IP chain. | The subject IP chain consistently outperformed the predicate in the 0–0.5 lp/mm range. |
| Low Contrast Detectability (LCD) | Subject IP chain performed significantly better than the predicate (p < 0.05), or no statistically significant difference in most test cases. | Across all conditions, the subject IP chain consistently demonstrated lower percent contrast values than the predicate (superior LCD performance), with improvements statistically significant in all cases (p < 0.05). |
| Contrast-to-Noise Ratio (CNR) of High Contrast Object | Subject IP chain performed significantly better than the predicate (p < 0.05), or no statistically significant difference in most test cases. | The subject IP chain significantly outperformed the predicate in all cases (p < 0.05), indicating a consistent and statistically significant improvement in CNR. |
| Hi-Def Mode Bench Test Results | ||
| Change in Image Level, Noise Magnitude and SNR | Image-level similarity using TOST, and noise magnitude and SNR properties equivalent to or better than the predicate (one-sided Student's t-test). | Noise and SNR properties of the subject device were better than those of the predicate. |
| Noise Power Spectrum (NPS) | Absence of unexpected distortions (e.g., spikes) and a reduction in noise at high spatial frequencies. | Both NPS curves were smooth and free of unexpected distortions. The subject IP chain exhibited lower noise at spatial frequencies at mid and high frequencies above 2 cycles/mm. |
| Noise Texture via Kurtosis | Subject IP chain's kurtosis being significantly closer to 3 than the predicate (p < 0.05) in most test cases. | The subject IP chain consistently met this criterion, indicating a more Gaussian-like noise texture and statistically lower kurtosis than the predicate. |
| Modulation Transfer Function (MTF) | Maintained or improved NEQ in the higher spatial frequency range (referencing Test 5). | Both MTF curves were smooth and free of unexpected distortions, and the subject IP chain demonstrated lower spatial resolution than the predicate chain (this needs to be read in conjunction with the NEQ results for success criteria). |
| Noise Equivalent Quanta (NEQ) | Subject IP chain exhibiting higher NEQ in the mid to high spatial frequency range compared with the predicate IP chain. | Results demonstrated that the subject IP chain consistently outperformed the predicate in mid and high frequencies. |
| Low Contrast Detectability (LCD) | Subject IP chain performed significantly better than the predicate (p < 0.05). | The results were considered acceptable, as the subject IP chain outperformed the predicate in the majority of ROI sizes (3 out of 4). |
| Contrast-to-Noise Ratio (CNR) of High Contrast Object | Subject IP chain performed significantly better than the predicate (p < 0.05) in most test cases. | Results showed that the subject IP chain significantly outperformed the predicate in all cases, indicating a consistent and statistically significant improvement in CNR. |
2. Sample size used for the test set and the data provenance
The document does not specify exact sample sizes for the test sets in terms of number of patients or images. The tests primarily utilized:
- Phantom data: Anthropomorphic chest phantoms and PMMA slab phantoms.
- Clinical datasets: Mentioned in image quality evaluations, but no further details provided regarding number or source.
- Data provenance: Not explicitly stated (e.g., country of origin, retrospective/prospective). The use of phantoms is a controlled laboratory setting.
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts
Not applicable. The reported tests are primarily quantitative bench tests using phantoms or objective image quality metrics, not dependent on expert interpretation for ground truth.
4. Adjudication method (e.g. 2+1, 3+1, none) for the test set
Not applicable, as the tests involve quantitative metrics measured from phantoms or images, rather than human expert adjudication.
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 MRMC comparative effectiveness study involving human readers is mentioned in the provided text. The testing focuses on objective image quality metrics using phantoms and clinical datasets.
6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done
Yes, the described performance testing is a standalone assessment of the αEvolve Imaging chain, including the Artificial Intelligence Denoising (AID) algorithm, without a human-in-the-loop component. The "reported device performance" directly reflects the algorithm's impact on image quality parameters.
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc)
The ground truth for most tests is derived from:
- Physical phantoms: Anthropomorphic chest phantoms and PMMA slab phantoms, which provide known geometries and material properties for objective measurement.
- Defined physical metrics: Metrics like image level, noise magnitude, SNR, NPS, kurtosis, MTF, NEQ, LCD, and CNR are calculated based on the image data and the known characteristics of the phantoms. There is no "ground truth" established by clinical experts or pathology in these technical bench tests.
8. The sample size for the training set
The document does not provide any information regarding the training set size for the Artificial Intelligence Denoising (AID) algorithm.
9. How the ground truth for the training set was established
The document does not provide any information on how the ground truth for the AID algorithm's training set was established.
Ask a specific question about this device
Page 1 of 1