Search Results

The HiSonic-TRD tinnitus masker is a wearable behindthe-ear style ultrasonic, bone conducted broadband noise and/or sweep frequency stimuli noise generator intended to generate noise of sufficient intensity and bandwidth to be used for tinnitus habituation therapy and is also suitable for tinnitus masking therapy. The HiSonic-TRD is intended to be used by those individuals who experience tinnitus, and do not need or desire amplification. The intended use of the HiSonic-TRD includes it's fitting by a qualified audiologist or other hearing healthcare professional familiar with the diagnosis of tinnitus and subsequent rehabilitation therapies. The target population is the adult population over 18 years of age. This product is not intended for use in children under the age of 18.

The HiSonic-TRD generates ultrasound energy and couples the energy to the mastoid bone in order to activate the auditory nervous system, and produce auditory sensations that mask tinnitus. The device consists of a generator/amplifier unit, a piezoelectric transducer, and a headband to hold the transducer firmly against the mastoid bone. A 7.2V Lithium Ion battery powers the device. The tranducer is a piezoelectric ceramic that transmits ultrasonic energy in a typical operating frequency bandwidth of 19.5kHz-25.8kHz to the user. The generator/amplifier unit is worn on the body and houses the lithium ion battery. In addition, the body worn unit has an "on/off" switch, a noise/sweep selector switch and a volume control wheel for use by the patient. Both broadband noise and sweep frequency stimuli are incorporated in the device as masking stimuli. The HiSonic-TRD generates and amplifies electronic oscillations, and then converts them into ultrasound energy with a piezoelectric ceramic transducer. The ultrasound energy is emitted as a broadband noise or as a sweep frequency.

Here's an analysis of the acceptance criteria and study information for the HiSonic-TRD based on the provided text:

No specific quantitative acceptance criteria relating to device performance metrics (e.g., sensitivity, specificity, accuracy, or a threshold for masking effectiveness) are explicitly stated in the document. The document primarily focuses on safety and demonstrating that the device is "substantially similar in effectiveness" to predicate devices.

However, based on the performance data presented, we can infer some desired outcomes which function as de facto acceptance criteria.

1. Table of Acceptance Criteria and Reported Device Performance

Study Details:

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

• Safety Study: 11 human subjects. Data provenance is prospective (study conducted specifically for this submission) in the USA (University at Buffalo Speech-Language and Hearing Clinic).
• Efficacy Study: 20 human subjects. Data provenance is prospective (study conducted specifically for this submission) in the USA (Oregon Health Sciences University Tinnitus Clinic).
• Bioacoustics Research Laboratory (Safety measurements): The text does not specify a "sample size" in terms of cases for these technical measurements, as it involves physical measurement of the device, not patient data in the same way. The provenance is the USA (University of Illinois Bioacoustics Research Laboratory).

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

• Safety Study: The text does not mention specific "experts" establishing ground truth in terms of diagnostic labels for a test set. The study observed changes in auditory thresholds, speech intelligibility, and adverse events. These were likely measured by qualified clinicians (e.g., audiologists) at the University at Buffalo Speech-Language and Hearing Clinic. The qualifications of these individuals are not specified beyond their affiliation with the clinic.
• Efficacy Study: The text does not explicitly detail "experts" creating a ground truth label for the test set. Instead, researchers at the Oregon Health Sciences University Tinnitus Clinic (implied to be experts familiar with tinnitus assessment) conducted baseline measurements and assessed the device's ability to mask tinnitus based on a "threshold approach to audibility" and measuring "masking effects." The selection criteria required subjects with "clinically significant tinnitus," implying a prior diagnosis by a qualified professional, but not an independent "ground truth" establishment for the study itself by experts.

4. Adjudication method for the test set:

• The text does not describe any formal adjudication method (e.g., 2+1, 3+1, none) for either the safety or efficacy studies. The measurements and observations appear to be direct outputs or assessments from the clinics.

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:

• No, a multi-reader multi-case (MRMC) comparative effectiveness study was not done. This device is a treatment device, not an AI diagnostic tool that assists human readers. The efficacy study compared the device's masking rate to that of "FDA cleared tinnitus maskers," which is a comparative effectiveness claim, but not in the context of human reader improvement with AI assistance.

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

• This question is not applicable. The HiSonic-TRD is a medical device for treatment, not an algorithm. The "standalone" performance here would be the device functioning on its own, which is what the studies assessed (its ability to generate sound and mask tinnitus irrespective of operator skill in using it).

7. The type of ground truth used:

• Safety Study: The ground truth was primarily based on objective measurements of audiological parameters (auditory thresholds, speech intelligibility) and self-reported adverse events. There isn't a "ground truth" label in the sense of a diagnosis being confirmed by an external gold standard.
• Efficacy Study: The ground truth for efficacy was the subjective experience of tinnitus and its masking. The study selected subjects whose "tinnitus was present for at least twelve months and that the tinnitus showed little variability in loudness, localization or sound quality." Masking effectiveness was assessed by subjects reporting audibility of the device's sound and the effect on their tinnitus, as measured by a clinical "threshold approach to audibility." This relies on patient self-report and clinical assessment rather than an objective "pathology" or external "outcomes data" in the typical sense for diagnostic devices.

8. The sample size for the training set:

• There is no mention of a "training set" as this device is not an AI/machine learning model. The studies described are for validation of the device itself.

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

• Not applicable, as there is no training set for this device.

Ask a specific question about this device

Intended to be used as a wearable sound amplifying device to compensate for mild to severe hearing loss.

Model WMR BTE is a dual-channel, wide-dynamic range input instrument which allows for loudness correction compression by changing compression ratios within the two bands. The WMR separates incoming signals into high and low frequency bands which can be independently compressed prior to recombination and amplification. The electronics for the hearing instrument are contained in a custom plastic case.

The Model WMR BTE features three adjustable trimmers that can only be adjusted by the dispenser to meet individual patient requirements. The user may adjust the gain (loudness) of the hearing instrument by rotating the volume control dial to his/her most comfortable listening level. This hearing instrument does not require a special fitting system; however, the trimmer controls may only be adjusted by an authorized hearing aid dispenser, NOT by the consumer. Notes to this effect are included in the User Instruction Booklet included in this submission.

• The crossover frequency between the low band and high band channels; 1)
• 2. High band compression ratio;
• 3. Low band compression ratio

This 510(k) summary describes a hearing aid, which is a medical device that predates the widespread use of software-driven AI or complex algorithms as we understand them today in healthcare. Therefore, the information requested in your prompt (especially concerning AI/ML performance, ground truth, expert adjudication, MRMC studies, and training/test sets) is largely not applicable to this document.

However, I can extract the information that is present regarding performance and criteria for this type of device.

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

For a hearing aid submitted in 1996, "acceptance criteria" and "reported device performance" are primarily tied to compliance with established industry standards rather than a quantitative accuracy metric common for modern AI devices.

2. Sample sized used for the test set and the data provenance (e.g. country of origin of the data, retrospective or prospective)

Not applicable. This device is a hardware product. The "test set" would be the device itself tested against an industry standard. There's no "data provenance" in the sense of patient data.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts (e.g. radiologist with 10 years of experience)

Not applicable. "Ground truth" in the context of this device refers to the physical and acoustical properties as defined by the ANSI standard. This is established by engineering and acoustical measurement, not expert clinical interpretation of data.

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

Not applicable. There's no "adjudication" of results in the sense of multiple human reviewers. Performance is based on objective, quantifiable physical measurements against a standard.

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

Not applicable. This is a standalone hearing aid, not an AI-assisted diagnostic or interpretive tool that would be used by "human readers."

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

Yes, in a way. The device itself (the "algorithm" in a loose interpretation of its function) is tested standalone against the ANSI standard. Its performance is measured directly without human interaction beyond operating the test equipment.

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

The "ground truth" is defined by the American National Standard Specification of Hearing Aid Characteristics, ANSI S3.22-1987. This standard sets objective, measurable criteria for hearing aid performance, such as frequency response, maximum output, and distortion.

8. The sample size for the training set

Not applicable. This is a hardware device. There is no "training set" in the machine learning sense. The device is designed and engineered.

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

Not applicable. No training set in the AI sense. The design of the device is based on established principles of acoustics and audiology, aiming to meet the performance criteria outlined in the ANSI standard for addressing hearing loss.

Ask a specific question about this device