(339 days)
The Neuspera Neurostimulation System (NNS) is indicated for pain management in adults who have severe intractable chronic pain of peripheral nerve origin, as the sole mitigating agent or as an adjunct to other modes of therapy used in a multidisciplinary approach. The system is not intended to treat pain in the craniofacial region.
The Neuspera Neurostimulation System (NNS) is also used for trial stimulation (no longer than 30 days) to determine efficacy before recommendation for a permanent (long term) implant.
The Neuspera Neurostimulation System is used for peripheral nerve stimulation to provide therapeutic relief for chronic, intractable pain of peripheral nerve origin. The System consists of an implantable pulse generator (IPG), electrode array, surgical implant tools, wireless worn transmitter, clinician programmer and a patient controller. The implantable pulse generator is a miniature implanted neurostimulator, powered by an externally worn wireless transmitter device which contains a rechargeable battery.
The provided text is a 510(k) summary for the Neuspera Neurostimulation System (NNS). It details the device's indications for use, its components, and a comparison to predicate devices, focusing on technological characteristics. However, this document describes a neurostimulation system for pain management, NOT an AI/ML medical device for image analysis or diagnosis.
Therefore, the information required to answer your prompt, which is specifically related to acceptance criteria and studies for AI/ML device performance (e.g., accuracy, sensitivity, specificity, expert ground truth, MRMC studies), is not present in this 510(k) summary. The summary focuses on hardware specifications, electrical properties, biocompatibility, and animal studies for an implantable medical device, and explicitly states "Clinical evaluation is not required".
To directly address your request, if this were an AI/ML device submission, here's what the answer would look like (hypothetically, based on typical AI/ML medical device FDA submissions):
Hypothetical Response (if this were an AI/ML device, assuming typical FDA AI/ML study requirements):
This 510(k) summary does not appear to be for an AI/ML medical device that requires clinical performance studies based on human reader performance or algorithm-only metrics. The device, Neuspera Neurostimulation System (NNS), is an implanted peripheral nerve stimulator for pain relief. The provided documentation focuses on engineering specifications, biocompatibility, and non-clinical testing (functional, performance, MRI testing, animal studies) to demonstrate substantial equivalence to a predicate device.
The summary explicitly states: "Clinical evaluation is not required for the Neuspera Neurostimulation System as the indications for use are equivalent to the legally marketed predicate device and referenced device. These types of devices, including versions of the legally marketed predicate device, have been on the market for many years with a proven safety and efficacy for the use of the device. Therefore, Neuspera determined that bench and non-clinical testing are sufficient to demonstrate that the Neuspera Neurostimulation System is as safe and effective as the predicate device."
Therefore, the requested information regarding acceptance criteria, performance metrics (like sensitivity, specificity), data provenance, expert ground truth establishment, adjudication methods, MRMC studies, or standalone algorithm performance, which are typical for AI/ML diagnostic or prognostic devices, is not applicable or available in this specific 510(k) submission for the Neuspera Neurostimulation System.
If this were an AI/ML device submission, the following sections would be populated (but cannot be from the provided text):
- Table of acceptance criteria and reported device performance:
(Hypothetical example for an AI/ML device)
| Metric | Acceptance Criteria | Reported Device Performance |
|---|---|---|
| Sensitivity | ≥ 90% | 92.5% |
| Specificity | ≥ 80% | 85.1% |
| AUC (ROC) | ≥ 0.90 | 0.93 |
| PPV | ≥ 75% | 78.2% |
| NPV | ≥ 95% | 96.8% |
-
Sample size used for the test set and the data provenance:
(Hypothetical example for an AI/ML device)- Test Set Sample Size: E.g., 500 cases (e.g., medical images).
- Data Provenance: Retrospective, collected from multiple institutions across the United States, Europe, and Asia.
-
Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
(Hypothetical example for an AI/ML device)- Number of Experts: E.g., 3 independent board-certified radiologists.
- Qualifications: Each radiologist had a minimum of 10 years of experience specializing in (e.g., thoracic imaging) and were blinded to the device's output.
-
Adjudication method for the test set:
(Hypothetical example for an AI/ML device)- Adjudication Method: 2+1; if two initial readers disagreed, a third senior expert (adjudicator) reviewed the case to establish the final ground truth.
-
If a Multi-Reader Multi-Case (MRMC) comparative effectiveness study was done:
(Hypothetical example for an AI/ML device)- MRMC Study: Yes, an MRMC study was conducted comparing human reader performance with and without AI assistance.
- Effect Size: Human readers demonstrated a statistically significant improvement in diagnostic accuracy (e.g., 15% increase in AUC) when assisted by the AI device compared to unassisted reading. The sensitivity increased by X% and specificity by Y%.
-
If a standalone (i.e., algorithm only without human-in-the-loop performance) was done:
(Hypothetical example for an AI/ML device)- Standalone Performance: Yes, standalone performance was evaluated on the test set. The algorithm achieved a sensitivity of 92.5% and a specificity of 85.1%.
-
The type of ground truth used:
(Hypothetical example for an AI/ML device)- Ground Truth Type: Expert consensus (from the expert radiologists) reviewed against relevant clinical outcomes data (e.g., biopsy results, surgical pathology, or patient follow-up data for disease progression/regression).
-
The sample size for the training set:
(Hypothetical example for an AI/ML device)- Training Set Sample Size: E.g., 10,000 cases.
-
How the ground truth for the training set was established:
(Hypothetical example for an AI/ML device)- Training Set Ground Truth: Established by a combination of clinical reports, a subset reviewed by a single board-certified radiologist, and confirmed with pathology results or long-term patient follow-up where available. Automated methods (e.g., natural language processing of reports) were also used for initial labeling, with a portion of cases undergoing expert review for quality control.
§ 882.5870 Implanted peripheral nerve stimulator for pain relief.
(a)
Identification. An implanted peripheral nerve stimulator for pain relief is a device that is used to stimulate electrically a peripheral nerve in a patient to relieve severe intractable pain. The stimulator consists of an implanted receiver with electrodes that are placed around a peripheral nerve and an external transmitter for transmitting the stimulating pulses across the patient's skin to the implanted receiver.(b)
Classification. Class II (performance standards).