Search Results

NeuroStar Advanced Therapy is indicated for the treatment of Major Depressive Disorder in adult patients who have failed to receive satisfactory improvement from prior antidepressant medication in the current episode.

The NeuroStar Advanced Therapy System is a computerized, electromechanical medical device that produces and delivers non-invasive, magnetic stimulation using brief duration rapidly alternating, or pulsed, magnetic fields to induce electrical currents directed at spatially discrete regions of the cerebral cortex. This method of cortical stimulation by application of brief magnetic pulses to the head is known as Transcranial Magnetic Stimulation (TMS).

NeuroStar Advanced Therapy is a non-invasive tool for the stimulation of cortical neurons for the treatment of adult patients with Major Depressive Disorder (MDD) who have failed to receive satisfactory improvement from prior antidepressant medication in the current episode. NeuroStar Advanced Therapy is used for patient treatment by prescription only under the supervision of a licensed physician and can be used in both inpatient and outpatient settings including physician's offices, clinics, and hospitals.

The proposed changes to the NeuroStar Advanced Therapy System described in this Traditional 510(k) Premarket Notification include the introduction of a new feature that allows the NeuroStar Advanced Therapy system to perform the TMS therapy known as intermittent theta burst stimulation (iTBS). NeuroBurst is the proprietary name for the iTBS treatment conducted by the NeuroStar Advanced Therapy System. The NeuroBurst treatment protocol consists of a burst of three (3) pulses at 50Hz with a 160ms interval between bursts. The protocol uses a train that consists of five (5) bursts per second for two (2) seconds with an eight (8) second interval between trains. A treatment session lasts for 20 trains or 3.3 minutes.

The NeuroStar Advanced Therapy System is an integrated system consisting of a combination of the following components:

• Mobile Console for housing the electronics and includes a software controlled graphical user interface, display monitor, display arm, and gantry that supports the treatment coil.
• Ferromagnetic Coil for delivering treatment.
• Head Support System for positioning the treatment coil and includes a laser-guided alignment system.
• Multi-use consumable SenStar Treatment Link for contact sensing of the treatment coil with the patient's head and magnetic field quality control.
• TrakStar Patient Data Management System for recording patient data and includes a stand-alone computer and data management software.

The provided text is a 510(k) summary for the NeuroStar Advanced Therapy System, primarily focusing on demonstrating substantial equivalence to a predicate device and introducing a new feature (iTBS treatment, or NeuroBurst). It lacks the specific details required to answer your questions about acceptance criteria and a study proving a device meets these criteria in the context of an AI/ML medical device.

The document describes a medical device (Transcranial Magnetic Stimulation System) for treating Major Depressive Disorder, not an AI/ML algorithm that predicts or diagnoses based on data. Therefore, the concepts of "test sets," "ground truth," "expert consensus," "MRMC studies," or "standalone algorithm performance" as typically applied to AI/ML device validation are not discussed.

The "performance data" mentioned here refers to:

• Physical performance of the device: magnetic field characteristics, output waveform, linear output level, electromagnetic compatibility (EMC), and electrical safety.
• Comparison to predicate/reference devices: demonstrating that the NeuroStar system's technological characteristics, indications for use, and principles of operation are substantially equivalent. The clinical effectiveness of iTBS itself is supported by the reference device's prior FDA clearance, not a new clinical study presented in this 510(k) summary for the NeuroStar device's NeuroBurst feature.

Therefore, I cannot extract the requested information as it is not present in the provided text.

To illustrate what kind of information would be needed to answer your questions, here's a hypothetical example relevant to an AI/ML device, assuming the NeuroStar was an AI/ML device (which it is not, based on this document):

Hypothetical Example (If NeuroStar were an AI/ML device):

Let's imagine the NeuroStar Advanced Therapy System had an AI component designed to predict patient response to TMS therapy based on patient characteristics and brain imaging data.

1. A table of acceptance criteria and the reported device performance

2. Sample size used for the test set and the data provenance

• Test Set Sample Size: 500 patients
• Data Provenance:
  • Country of Origin: Multi-center study including data from the United States (40%), United Kingdom (30%), and Canada (30%).
  • Retrospective/Prospective: Data collected prospectively from an observational cohort study specifically designed for AI model validation.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts

• Number of Experts: 3 independent expert psychiatrists.
• Qualifications of Experts: Each psychiatrist had at least 15 years of clinical experience in treating Major Depressive Disorder and specialized in TMS therapy. They were board-certified in psychiatry in their respective countries.

4. Adjudication method (e.g. 2+1, 3+1, none) for the test set

• Adjudication Method: 3-reader consensus. If all three experts agreed on the patient's responder status (responder or non-responder), that was taken as the ground truth. In cases of disagreement (e.g., 2 agree, 1 disagrees), a fourth, senior adjudicating psychiatrist (not involved in initial readings) reviewed the case and made the final determination.

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

• MRMC Study: Yes, an MRMC study was conducted.
• Effect Size:
  • Without AI Assistance: Average AUC for human readers was 0.70 (95% CI: 0.68-0.72).
  • With AI Assistance: Average AUC for human readers improved to 0.80 (95% CI: 0.78-0.82).
  • Improvement (Effect Size): The AI assistance led to an average increase of 0.10 in AUC for human readers (p < 0.001), indicating a statistically significant improvement in their ability to predict treatment response.

6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done

• Standalone Performance: Yes, standalone performance was evaluated. The AI algorithm achieved an AUC of 0.88, Sensitivity of 82%, and Specificity of 75%.

7. The type of ground truth used (expert consensus, pathology, outcomes data, etc)

• Type of Ground Truth: Clinical outcomes data (validated depression symptom scales, e.g., MADRS or HAM-D, post-treatment scores indicating significant reduction in symptoms) independently assessed by the 3 expert psychiatrists via patient records and follow-up data abstraction. This was augmented by their consensus-based clinical determination of "responder" status.

8. The sample size for the training set

• Training Set Sample Size: 5,000 patients.

9. How the ground truth for the training set was established

• Ground Truth Establishment for Training Set: Ground truth for the training set was established through retrospective review of patient electronic medical records from multiple institutions. "Responder" status was defined by a pre-specified threshold reduction in a standardized depression symptom scale (e.g., ≥50% reduction in HAM-D scores) at the end of the TMS treatment course, as documented by treating physicians. Data was then cross-referenced and anonymized by research coordinators. While not as rigorously adjudicated as the test set, standard clinical documentation was considered sufficient for training purposes given the large volume of data.

Ask a specific question about this device