Search Results
Found 2 results
510(k) Data Aggregation
(179 days)
Models IRIS 20 or IRIS 41 SSXI detector combined with IRIS-View Image Acquisition control, display viewing, and archiving console is indicated for use as a SSXI Portal Imaging Device used in conjunction with Medical Accelerator Radiotherapy Devices as an alternative to Conventional Portal Film.
The IRIS flat panel X-ray detector is available in two sizes, 20 x 20 cm (IRIS 20) and 41 x 41 cm (IRIS 41), with an Element Matrix of 256 x 256 and 1024 x 1024 respectively. X-rays are converted to light using a Gd202S(Tb) opto-mechanical luminescent screen coupled to an Amorphous Silicon Thin Transistor Photodiode Integrating Storage Matrix Array. The IRIS-View console accepts the digital signals from the detectors through a PC card interface into a standard PC with a Microsoft Windows operating system. Software processing of the data is applied to enhance the visibility of the viewed image seen on a standard PC Color Monitor.
The provided text describes the Bio-Scan IRIS Solid State X-ray Imager, a device intended for use as a Solid State X-ray Imager (SSXI) Portal Imaging Device in conjunction with Medical Accelerator Radiotherapy Devices, serving as an alternative to Conventional Portal Film.
Here's an analysis of the acceptance criteria and the study that proves the device meets them:
1. Table of Acceptance Criteria and Reported Device Performance
The submission does not explicitly define quantitative acceptance criteria for clinical effectiveness in a pass/fail manner. Instead, the effectiveness study focuses on comparing the quality of IRIS images to conventional X-ray film images, with the implicit goal being that IRIS images are at least equivalent, if not superior, to film. Regulatory requirements are met through substantial equivalence to a predicate device and adherence to various standards (ISO, CEI, EN).
However, based on the nonclinical and clinical considerations, we can infer some key performance aspects and their reported values.
| Acceptance Criteria (Inferred) | Reported Device Performance |
|---|---|
| Nonclinical Performance (Technical Specifications) | |
| Active Area (IRIS 20) | 192 x 192 mm |
| Active Area (IRIS 41) | 409.6 x 409.6 mm |
| Matrix (IRIS 20) | 256 x 256 (750 x 750 µm) |
| Matrix (IRIS 41) | 1024 x 1024 (400 x 400 µm) |
| Fill Factor | 80% |
| Maximum Energy Detection | Up to 30 MeV accelerator beam source |
| Measured Linear X-ray Range | From 50kV to 117kV |
| Detector Materials | Gd2O2S(Tb) opto-mechanical luminescent screen; Amorphous Silicon Thin Film Transistor Photodiode Integrating Storage Matrix Array |
| Data Output Format | DICOM III standard format (Bitmap also available) |
| DQE (Detective Quantum Efficiency) | > 65% at 0.3 mm^-1 frequency for 10 µGy exposure |
| SNR (Signal-to-Noise Ratio) | Proportional to square root of exposure; SNR vs. Entrance Kerma measured from 0 to 25 µGy for 50kV to 117kV, showing SNR from 25 to 275 |
| MTF (Modulation Transfer Function) | Measured from 0.1 to 1.275 mm^-1 frequency, showing MTF from 0.98 to 0.01 |
| Aliasing | Matrix doubling algorithm utilized to eliminate aliasing |
| Dynamic Range | Slice Profile of a 45° wedge filter image shows input energy from 18MeV vs. 16 bit digital intensity output. Dose Response plotted from 0 µGy to 40 µGy for 50 to 117 kV, showing linear response from 0 to 65,535 intensity units. |
| Lag Time | 35 µsec signal charge transfer and reset to zero; 280 msec total before new integration |
| Pixel Defects | Up to 5% may be noisy or inefficient, corrected by adjacent pixel averaging algorithm and calibration protocol. |
| Device Ready Time | One hour after power turn-on; Can acquire a new image 280 msec after an exposure. |
| Dose Requirement (for Acceptable SNR) | Acceptable SNR of 100 achieved at Entrance Air Kerma of 2 µGy from 50 to 117kV |
| Stability | Stable after one hour of power on. |
| Uniformity | Weekly QA protocol produces a calibration matrix which corrects for non-uniformity of gain of individual pixels and X-ray beam non-uniformities. |
| Frame Rate (Reuse Rate for static images) | 4 fps to 10 fps depending on field of view selected (not for fluoroscopic applications). |
| Clinical Effectiveness | |
| Image Quality compared to X-ray film (Implicit Acceptance: Equivalent or better) | Better: 27 out of 30 comparison images (90%) rated better than X-ray film. Equivalent: 3 out of 30 comparison images (10%) rated equivalent to X-ray film. Poorer: In no situation was any quality of the Bio-Scan Image rated of poorer quality than the X-ray film image. |
| Compliance with Regulatory Standards | Met ISO 9001, ISO 13585/EN 46001, EN 1441, EN1041, CEI 61223-3-1, CEI 61223-3-3, EN 980, EN 540, EN 60601 (FCC Class A), EN 60950-A4. Certified by SGS Yarsley International Certification Services Ltd and INTERTest Systems. |
2. Sample Size Used for the Test Set and Data Provenance
- Sample Size: 30 patients were included in the clinical trial. For each patient, two images were compared (one X-ray film and one IRIS SSXI image), leading to a total of 30 comparison images (60 images in total: 30 film + 30 IRIS).
- Data Provenance: The study was sponsored and conducted by the University of Pennsylvania Radiation Oncology Department, Division of Medical Physics. This suggests the data originated from the United States. The study was a prospective clinical trial, carried out under Local IRB approval, with patients scheduled to undergo radiation therapy.
3. Number of Experts and Qualifications for Ground Truth
- Number of Experts: 7 physicians were involved in reading and comparing the images.
- Qualifications: "Specialists in the radiation treatment of the specific target organ lesions to be irradiated" and "Physicians with expertise in the Speciality of Radiation Oncology for the target organ." Specific years of experience or board certifications are not provided.
4. Adjudication Method for the Test Set
The document mentions that "Reading was done by specialists... A table of reader responses was developed comparing the results..." It then states that "three out of the total of 30 image comparisons by 7 physicians, the rating was considered equivalent." This implies that the 7 physicians individually rated each comparison. It does not explicitly state an adjudication method (like 2+1 or 3+1 consensus). It sounds more like individual ratings were collected, and then a summary of these individual ratings was compiled to describe the overall outcome (e.g., how many times IRIS was rated better/equivalent/worse by the group across all images). Without further detail, we cannot confirm a formal adjudication process beyond individual expert evaluation.
5. Multi-Reader, Multi-Case (MRMC) Comparative Effectiveness Study
- Was it done? Yes, a comparative effectiveness study was conducted where 7 physicians compared IRIS images to conventional X-ray film images for 30 patients.
- Effect Size of Human Readers Improvement with AI vs. without AI assistance: This study was not a "human reader with AI vs. human reader without AI" study. It was a direct comparison of the image quality from the new device (IRIS) against the existing standard (X-ray film), as interpreted by human readers. The readers were assessing the image quality produced by two different imaging modalities, not the performance of an AI assisting a human reader. Therefore, the effect size of how much human readers improve with AI vs. without AI assistance is not applicable here. The study concluded that for 90% of cases, IRIS images were rated better by the physicians than X-ray film, and for 10% of cases, they were equivalent.
6. Standalone (Algorithm Only) Performance
- Was it done? No, the device is an imaging system (hardware and associated software for image acquisition, viewing, and archiving), not an AI algorithm designed to interpret images independently. The effectiveness study involved human physicians interpreting the images produced by the device. Therefore, a standalone (algorithm only) performance assessment, as understood for AI systems, was not performed or relevant in this context.
7. Type of Ground Truth Used
The ground truth for the clinical effectiveness study was established through expert consensus/opinion (or rather, expert comparison and rating). The radiologists/oncologists directly compared the quality of the IRIS images against the X-ray film images, which was the existing standard of care. There was no mention of an independent, definitive "ground truth" like pathology for image interpretation accuracy, but rather a direct comparison of imaging modalities based on visual assessment of image quality for the purpose of radiation treatment.
8. Sample Size for the Training Set
The document does not mention a separate "training set" in the context of machine learning. The IRIS device is a digital X-ray imager, and its image processing is described in terms of "Software processing of the data is applied to enhance the visibility of the viewed image." This suggests traditional image processing algorithms rather than a machine learning model that requires a distinct training dataset. If such an algorithm exists internally, its training set is not described.
9. How Ground Truth for the Training Set Was Established
As no training set (in the machine learning sense) is explicitly mentioned, the method for establishing its ground truth is not applicable/not provided. The device's characteristics and performance are described through physical testing, regulatory compliance, and a reader study comparing image output.
Ask a specific question about this device
(349 days)
Ask a specific question about this device
Page 1 of 1