(248 days)
Mini-X is intended for use by qualified/trained medical professionals who fully understand the safety information, emergency procedures, and the device's capabilities and function. The device provides fluoroscopic imaging and is used for guidance and visualization during diagnostic radiography and surgical procedures of the extremities. The device will be used in healthcare facilities inside and outside the hospital, using various methods for the extremities on all patients except neonates (birth to one month) within the limits of the device. Applications can be performed with the patient sitting, standing, or lying in a prone or supine position. The system is not intended for mammography applications. (Rx Only)
The Mini-X system, a unique mobile imaging system, can acquire, process, and display fluoroscopic images. Its portability allows for easy positioning within a room and movement from room to room within a facility, facilitating on-demand fluoroscopic examinations. The system's innovative design incorporates a low-powered mono-block generator and a dynamic flat-panel detector, enabling it to be powered through a single-phase 120VAC power outlet.
The Insight Enhanced™ DRF Digital Imaging System, a cutting-edge tool for healthcare professionals, offers full control over the imaging chain. It empowers the operator to view and enhance high-definition fluoroscopy images up to 30 fps, bringing out diagnostic details that are challenging or impossible to see using conventional imaging techniques. The system's versatility is demonstrated by its ability to store images locally for short-term storage, produce hardcopy images with a laser printer, or send images over a network for longer-term storage. Its primary components, including a dynamic flat panel detector, monitors, and an image processor PC, underscore its comprehensive and advanced capabilities.
It seems there's a misunderstanding of the provided text. The document is an FDA 510(k) Clearance Letter for a mobile digital imaging system. It describes the device, its intended use, and its substantial equivalence to predicate devices based on technological similarities and differences, and outlines general testing performed (bench testing, compliance with IEC standards).
**Crucially, this document does not describe:
- An AI/ML device or its performance characteristics.
- Acceptance criteria related to AI/ML performance metrics (e.g., sensitivity, specificity, AUC).
- A study proving an AI device meets acceptance criteria.**
The document discusses imaging quality and radiation dose comparisons, and compliance with general medical device standards. It briefly mentions "software" and "cybersecurity" in generic terms, but not in the context of an AI algorithm or its performance.
Therefore, I cannot fulfill the request to describe the acceptance criteria and a study proving an AI device meets those criteria based on the provided text, as the text describes a standard imaging system, not an AI-powered one.
To answer your request, I would need a 510(k) submission document or a scientific paper that specifically details the clinical validation of an AI/ML medical device, including the elements you listed.
Hypothetical Example (if the document were about an AI device):
If this document were to refer to an AI device for, say, detecting fractures on Mini-X images, the relevant sections would typically be under "Non-Clinical and/or Clinical Tests Summary & Conclusions" (Page 7) or a dedicated "Performance Benchmarks" section, which are absent here for AI.
To illustrate what you're asking for, here's an example of how I would answer if the provided text contained the necessary information about an AI device:
(This is a hypothetical response, as the provided text doesn't contain this information.)
Based on the provided (hypothetical) information regarding an AI-powered diagnostic device, here's a description of the acceptance criteria and the study proving the device meets them:
1. Table of Acceptance Criteria and Reported Device Performance (Hypothetical)
| Performance Metric | Acceptance Criteria | Reported Device Performance |
|---|---|---|
| Sensitivity | ≥ 90% for detecting [condition A] | 92.5% |
| Specificity | ≥ 80% for detecting [condition A] | 85.1% |
| AUC | ≥ 0.90 for [condition A] detection | 0.93 |
| False Positive Rate | ≤ 0.05 cases/image for [normal anatomy] | 0.03 cases/image |
| Inference Time | ≤ 2 seconds per image for standard processing | 1.5 seconds/image |
2. Sample Size Used for the Test Set and Data Provenance (Hypothetical)
- Test Set Sample Size: 1500 unique imaging studies.
- Data Provenance: Retrospective and prospective data collected from multiple hospitals across the United States (70% retrospective, 30% prospective). The retrospective data covered a period of 5 years (2018-2023).
3. Number of Experts Used to Establish Ground Truth for the Test Set and Qualifications (Hypothetical)
- Number of Experts: A panel of 3 independent radiologists.
- Qualifications: All radiologists were board-certified with a minimum of 10 years of experience in diagnostic radiography, specializing in musculoskeletal imaging. One radiologist had subspecialty fellowship training in advanced imaging.
4. Adjudication Method for the Test Set (Hypothetical)
- Adjudication Method: 2+1 adjudication method was employed.
- Initially, two radiologists independently reviewed each case.
- If their interpretations agreed, that consensus was taken as the preliminary ground truth.
- If their interpretations disagreed, a third, senior radiologist served as an adjudicator and made the final decision to establish the ground truth.
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study (Hypothetical)
- MRMC Study Done: Yes, an MRMC study was conducted to evaluate the impact of AI assistance on human reader performance.
- Effect Size: The study demonstrated a significant improvement in reader performance. Human readers, when assisted by the AI device, showed an average 15% increase in sensitivity for detecting [condition A] and a 5% reduction in reading time per case, compared to reading without AI assistance, while maintaining specificity. The estimated Area Under the Free-Response Receiver Operating Characteristic (FROC) curve, a common metric in MRMC studies, improved from 0.78 (unaided) to 0.86 (AI-aided).
6. Standalone (Algorithm Only) Performance Study (Hypothetical)
- Standalone Study Done: Yes, a standalone performance evaluation was conducted on the full test set (1500 cases) against the established ground truth.
- Standalone Performance Metrics:
- Sensitivity: 92.5%
- Specificity: 85.1%
- F1-score: 0.88
- AUC: 0.93
7. Type of Ground Truth Used (Hypothetical)
- Type of Ground Truth: Expert consensus, established through the 2+1 adjudication process involving three qualified radiologists. In cases where available and relevant, this was supplemented or confirmed by pathology reports or follow-up outcomes data (e.g., surgical confirmation or clinical progression documented over 6 months).
8. Sample Size for the Training Set (Hypothetical)
- Training Set Sample Size: 50,000 imaging studies, collected from a diverse patient population.
9. How Ground Truth for the Training Set Was Established (Hypothetical)
- Ground Truth Establishment for Training Set: The ground truth for the training set was primarily established through a combination of:
- Radiologist Consensus: A larger team of 10 radiologists (separate from the test set readers) annotated the training data. Each image was reviewed by at least two radiologists, with disagreements resolved by an internal consensus committee.
- Clinical Records & Reports: For a subset of cases, ground truth was derived from detailed clinical reports, electronic health records, and existing radiology reports.
- Automated Labeling (with verification): For a large portion of the normal or clearly pathological cases, a pre-existing, highly accurate internal model was used for initial labeling, which was then systematically reviewed and corrected by human annotators to ensure high fidelity. All ambiguous or complex cases were subjected to full manual review by multiple radiologists.
§ 892.1650 Image-intensified fluoroscopic x-ray system.
(a)
Identification. An image-intensified fluoroscopic x-ray system is a device intended to visualize anatomical structures by converting a pattern of x-radiation into a visible image through electronic amplification. This generic type of device may include signal analysis and display equipment, patient and equipment supports, component parts, and accessories.(b)
Classification. Class II (special controls). An anthrogram tray or radiology dental tray intended for use with an image-intensified fluoroscopic x-ray system only is exempt from the premarket notification procedures in subpart E of part 807 of this chapter subject to the limitations in § 892.9. In addition, when intended as an accessory to the device described in paragraph (a) of this section, the fluoroscopic compression device is exempt from the premarket notification procedures in subpart E of part 807 of this chapter subject to the limitations in § 892.9.