Search Results

The NeuroStar Advanced Therapy System is a transcranial magnetic stimulation device. Specifically, it is a computerized, electromechanical medical device that produces and delivers non-invasive magnetic fields to induce electrical currents targeting specific regions of the cerebral cortex. Transcranial magnetic stimulation (TMS) is a non-invasive technique used to apply brief magnetic pulses to the brain. The pulses are administered by passing high currents through an electromagnetic coil placed adjacent to a patient's scalp. The pulses induce an electric field in the underlying brain tissue. When the induced field is above a certain threshold and is directed in an appropriate orientation relative the brain's neuronal pathway, localized axonal depolarizations are produced, thus activating neurons in the targeted brain region.

The NeuroStar System consists of a combination of hardware, disposable, and consumable supplies, which are required for the operation of the system. The basic configuration includes the following components:

Mobile Console
System Software
Treatment Chair
Head Support System
MT Cap
. D-Tect MT Accessory
TrakStar Data Management

AI/ML Overview

The NeuroStar Advanced Therapy System is indicated as an adjunct for the treatment of Major Depressive Disorder (MDD) in adolescent patients (age 15-21).

1. Table of Acceptance Criteria and Reported Device Performance

Acceptance Criteria (Success Criteria)	Reported Device Performance
A minimum of 50% of the primary per protocol analysis population meet the Individual Subject Success Criteria (defined as a meaningful reduction in depression symptoms based on PHQ-9 scores over a pre-post TMS treatment interval of 6 weeks and achieving remission of MDD symptoms).	77.8% (95% CI: 72.8%, 83.0%) of the primary per protocol population met the criteria.
The lower limit of the 95% confidence interval for individual success rate should exceed the pre-established Overall Study Success Criteria of a minimum 50%.	The lower limit (72.8%) exceeded 50% by 22.79%.
Proportion of responders (defined as a meaningful reduction in depression symptoms based on PHQ-9 scores) is statistically significant.	This proportion was found to be statistically significant at p < 0.0001.
Mean change in PHQ-9 scores from baseline to endpoint is statistically significant.	The mean change in PHQ-9 scores from baseline to endpoint was -10.0 ± 6.6 (indicating improvement).
30% or more patients attain remission of MDD symptoms (defined as post-treatment PHQ-9 of <5).	30.0% attained remission of MDD symptoms.

2. Sample Size Used for the Test Set and Data Provenance

Sample Size (Test Set): 1,169 per protocol adolescent patients (age 12-21) were included in the real-world data analysis.
Data Provenance: The data was retrospective real-world data (RWD) derived from the TrakStar registry, collected from patients across 347 TMS centers in the US over a span of 15 years, beginning in 2008.

3. Number of Experts Used to Establish Ground Truth and Qualifications

The document does not explicitly state the number of experts used to establish ground truth or their qualifications for the test set derived from the TrakStar registry. The clinical data was based on real-world data (RWD) which would imply physician diagnoses and assessments during routine clinical care, not necessarily a separate expert panel for ground truth within the study itself.

4. Adjudication Method for the Test Set

The document does not specify an adjudication method like 2+1 or 3+1 for the test set. The data was collected from clinical practice (TrakStar registry) and analyzed retrospectively.

5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study

No multi-reader multi-case (MRMC) comparative effectiveness study was mentioned where human readers improved with AI vs. without AI assistance. The study focuses on the effectiveness of the NeuroStar Advanced Therapy System itself, not on an AI-assisted diagnostic tool for human readers.

6. Standalone (Algorithm Only) Performance Study

The NeuroStar Advanced Therapy System is a medical device that delivers transcranial magnetic stimulation, not a standalone algorithm. Therefore, a standalone (algorithm only) performance study as typically understood for AI/ML devices, was not performed. The performance data relates to the device's clinical efficacy in treating MDD.

7. Type of Ground Truth Used

The ground truth for the real-world data analysis was based on clinical diagnosis of Major Depressive Disorder (MDD) according to DSM-4/ICD-9 or DSM-5/ICD-19 criteria, and Patient Health Questionnaire-9 (PHQ-9) scores for depression severity and remission, collected as part of routine clinical care and stored in the TrakStar registry.

8. Sample Size for the Training Set

The document does not explicitly mention a "training set" in the context of an AI/ML algorithm that requires training. The device itself is a non-AI medical device; however, the submission uses real-world clinical data and literature reviews to support its expanded indication. If "training set" refers to the data used to initially establish the device's efficacy, the previous FDA clearances (K083538, K130233, etc.) would have involved their own clinical studies. For this specific expanded indication, the submission relies on the large-scale retrospective analysis of 1,169 adolescent patients (treated with the NeuroStar device in real-world settings) and a literature review of 14 studies involving a total of 1,812 adolescents.

9. How the Ground Truth for the Training Set Was Established

As noted in point 8, this is not an AI/ML algorithm requiring a "training set" in the typical sense. The "ground truth" for the real-world data analysis and literature review, which serve as evidence for device efficacy, was established through:

Clinical Diagnoses: Primary diagnosis of MDD according to DSM-4/ICD-9 or DSM-5/ICD-19 criteria.
Validated Clinical Scales: Assessment using standardized and validated depression scales such as PHQ-9, HAMD-24, HAM-D, CDRS-R, QIDS-A17-SR, CGI-S, HDRS-17, BDI-II, MADRS, BPRS-C, and CGAS. These scores represent objective measures of symptom severity and improvement, as reported in clinical practice and research studies.
Physician Assessment: The inclusion criteria for the RWD study specified "Primary diagnosis of Major Depressive Disorder (MDD), according to DSM-4/ICD-9 or DSM-5/ICD-19 criteria applicable on the date treatment with NeuroStar Advanced Therapy begins," indicating physician-based diagnosis.

Ask a Question

Ask a specific question about this device

K Number

K220127

Device Name

NeuroStar TMS Therapy System, NeuroStar Advanced Therapy System, NeuroStar, NeuroStar Advanced Therapy for Mental Health

Manufacturer

Neuronetics, Inc

Date Cleared

2022-07-15

(178 days)

Product Code

Regulation Number

Type

Panel

Reference & Predicate Devices

K201158,K161519,K160703,K133408,K130233,K083538,K210201

Predicate For

K222230,K223154,K230029,K233621

Intended Use

The NeuroStar Advanced Therapy System is indicated for the treatment of depressive episodes and for decreasing anxiety symptoms for those who may exhibit comorbid anxiety symptoms in adult patients suffering from Major Depressive Disorder (MDD) and who failed to achieve satisfactory improvement from previous antidepressant medication treatment in the current episode.

Device Description

Mobile Console
System Software
Treatment Chair
Head Support System
TrakStar PC
TrakStar Software

AI/ML Overview

The NeuroStar Advanced Therapy System is indicated for treating depressive episodes and decreasing anxiety symptoms in adult patients with Major Depressive Disorder (MDD) who have not achieved satisfactory improvement from previous antidepressant medication.

Here's an analysis of the acceptance criteria and supporting studies:

Table of Acceptance Criteria and Reported Device Performance

Acceptance Criteria (Outcome Measures)	Reported NeuroStar Advanced Therapy System Performance
For Depression (Derived from Predicate Device Clearance, not explicitly detailed here for criteria beyond initial clearance)	O'Reardon et al., 2008 & George et al., 2010 (original clearance studies for MDD):
Response Rate (≥ 50% decrease in end score relative to baseline) for HAMD-17 and HAMD-24	Statistically significant improvement (p<0.05) in response rate for both HAMD-17 and HAMD-24 compared to sham.
	TrakStar (unpublished) RWD study (MDD):
Response Rate (≥ 6 point decrease in PHQ-9 score at end of treatment relative to baseline)	58.9% of patients achieved response.
Remission Rate (PHQ-9 < 5 at study endpoint)	30.4% of patients attained remission.
Mean Change in PHQ-9 scores from baseline to endpoint	-10.5 ± 6.7 (p<0.0001).
For Comorbid Anxiety Symptoms (Current Clearance Justification)	O'Reardon et al., 2008, George et al., 2010 (RCTs):
Statistically significant reduction in anxiety symptoms as measured by HAMD – A/S F (a priori defined secondary endpoint) between-group difference in mean change from baseline to 6-week endpoint. (p<0.05)	Statistically significant improvement (p<0.05) in mean change in HAMD – A/S F scores from baseline to 6-week endpoint, in favor of the active treatment group (p=0.023 for O'Reardon; p<0.05 for George et al. stated parenthetically in text, though table only has p<0.05).
Effect size (Cohen's d) for decreasing anxiety symptoms compared to Sham control.	0.36. This compares favorably to predicate device (Brainsway Deep TMS) effect sizes (0.34 and 0.36) for similar indications.
	TrakStar (unpublished) RWD study (Anxiety):
Response Rate (≥ 6 point improvement (reduction) in GAD-7 score at end of treatment relative to baseline)	65.5% of patients met the primary outcome measure for anxiety.
Remission Rate (GAD-7 < 5 at study endpoint)	34.0% of patients attained remission.
Mean Change in GAD-7 scores from baseline to endpoint	-8.0 ± 5.7 (p<0.0001).
Effect size (Hedges g)	-1.4.
	Tuinstra et al., 2022 (Retrospective analysis for Anxiety):
Remission Rate (HAMD AS/F < 7 or GAD-7 < 5)	HAMD AS/F < 7: 87.5%; GAD-7 < 5: 22.6%.
Response Rate (≥ 50% decrease in end score relative to baseline for HAMD AS/F or GAD-7)	HAMD AS/F: 50%; GAD-7: 41.5%.
Clinically Significant Change (final score ≤ 3.2 for HAMD AS/F; mean change from baseline to endpoint ≥ 6 for GAD-7)	HAMD AS/F: 37.5%; GAD-7: 41.5%.
Mean Change in scores from baseline to endpoint (6 weeks) for HAMD AS/F and GAD-7	HAMD AS/F: -3.53, p<0.001; GAD-7: -5.32, p<0.001.
Effect size (Hedges g)	-0.777.

Sample sizes used for the test set and data provenance:
The clinical performance data supporting the expanded indications primarily relies on two Randomized Controlled Trials (RCTs) and supportive real-world data (RWD) and observational studies.
- RCTs (Test Set):
  - O'Reardon et al. (2007/2008): N=301 patients (Active: 155, Sham: 146). Data provenance is from a clinical trial that evaluated the safety and efficacy of the device for MDD.
  - George et al. (2010): N=190 patients (Active: 92, Sham: 98). Data provenance is from a company-independent, NIMH-funded trial.
- Supportive Data (Test Set/Observational):
  - TrakStar (2022) study (unpublished): N=664 patients. This is a large-scale retrospective analysis of real-world data (RWD) derived from the TrakStar registry data, collected from patients across 75 TMS centers in the US over 13 years (since 2008). This is retrospective data.
  - Tuinstra et al. (2022): N=77 patients (57 with clinically significant anxiety symptoms). This is a retrospective analysis of medical records from patients with MDD and comorbid anxiety symptoms, over 3.5 years.
The provenance is primarily retrospective clinical trial data (for the original MDD indication, now leveraged for anxiety) and retrospective real-world data/medical chart reviews for the updated anxiety indication. The country of origin for the studies is not explicitly stated for all, but given the US FDA submission and typical clinical trial conduct for device clearances, it's highly probable to be primarily US-based, especially for the TrakStar registry.
Number of experts used to establish the ground truth for the test set and qualifications of those experts:
The studies described are clinical trials and retrospective analyses of patient outcomes. "Ground truth" in this context refers to the documented clinical status of patients (e.g., diagnosis of MDD, anxiety symptoms, response to treatment, remission). While clinical diagnoses and symptom assessments would typically be performed by qualified clinicians (e.g., psychiatrists, physicians), the specific number of experts involved in establishing the initial diagnoses or ground truth for each patient in the test sets, or their specific years of experience, is not specified in the provided text. The studies utilized standardized psychiatric rating scales (HAMD-17, HAMD-24, HAMD – A/S F, PHQ-9, GAD-7), suggesting these were administered and interpreted by trained personnel.
Adjudication method for the test set:
The provided text does not explicitly detail any adjudication method (e.g., 2+1, 3+1, none) for establishing ground truth or evaluating patient outcomes in the test sets. The RCTs and observational studies would have relied on standard clinical assessment procedures and the reporting of scores from validated psychiatric rating scales.
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, an MRMC comparative effectiveness study was not done. The NeuroStar Advanced Therapy System is a Transcranial Magnetic Stimulation (TMS) device, which is a therapeutic device targeting brain regions for treatment, not an AI-assisted diagnostic tool that requires human readers to interpret images or data. Therefore, the concept of "human readers improve with AI vs without AI assistance" does not apply to this device.
If a standalone (i.e., algorithm only without human-in-the-loop performance) was done:
No, a standalone (algorithm only) performance study was not done. The device is a therapeutic system that is administered by clinicians to patients. Its performance is evaluated through clinical trials and real-world usage in human subjects, not as an isolated algorithm.
The type of ground truth used:
The ground truth used for these studies is clinical outcomes data and expert-evaluated patient status according to standardized psychiatric rating scales.
- Diagnosis of Major Depressive Disorder (MDD), confirmed by clinicians.
- Level of depressive symptoms measured by scales like HAMD-17, HAMD-24, and PHQ-9.
- Level of anxiety symptoms measured by scales like HAMD – A/S F and GAD-7.
- Response to treatment (e.g., ≥50% reduction in symptom scores, or ≥6 point reduction for GAD-7/PHQ-9).
- Remission (e.g., GAD-7 < 5, PHQ-9 < 5, HAMD AS/F < 7).
The sample size for the training set:
The provided information describes clinical studies primarily for validation or performance demonstration of the device for specific indications. It does not mention a "training set" in the context of machine learning model development. The NeuroStar Advanced Therapy System is a physical medical device, not an AI/ML algorithm whose parameters are iteratively "trained" on data. The clinical data presented is for demonstrating safety and effectiveness.
How the ground truth for the training set was established:
As no explicit "training set" in the AI/ML sense is mentioned for this device, this question is not directly applicable. If interpreting "training set" loosely as the cumulative clinical experience and data that informed the device's development or initial indications, the "ground truth" would have been established through a combination of medical understanding of MDD, clinical research, previous clinical trials, and regulatory requirements for therapeutic interventions.

Ask a Question

Ask a specific question about this device

Page 1 of 1