Search Results

The Lumen® 155-SF self-fitting, wireless air conduction hearing aids are intended to amplify sound for individuals 18 years of age or older with perceived mild to moderate hearing impairment. They are adjusted by the user to meet the user's hearing needs. The device is intended for over-the-counter sales without the assistance of a hearing health care professional.

Per 21 CFR 874.3325 a self-fitting wireless air conduction hearing aid is a wearable sound amplifying device that is intended to compensate for impaired hearing and incorporates technology, including software, that allows users to program their hearing aids. This technology integrates user input with a self-fitting strategy and enables users to independently derive and customize their hearing aid fitting and settings. Self-fitting wireless air conduction hearing aids are class II medical devices.

The Lumen 155-SF self-fitting hearing aid(s) with Sentibo application is a self-fitting wireless air conduction Behind-The-Ear (BTE) hearing aid consisting of the Intricon hardware and Sentibo software to be used on an Apple mobile device (mobile device not included) including wired lightning connector Apple EarPods for the assessment (not included) designed for a single user. The Sentibo app is available only on iOS. The Lumen 155-SF self-fitting OTC hearing aid(s) with Sentibo application is a behind the ear device that includes self-adjustable coupling by means of a slim tube and ear tip/open dome. The hearing aids can be fine-tuned by the user or remotely at the request of the user via the use of QR codes. Once the hearing aids are switched on, the software (Sentibo app) is required for the initial set-up, which is done via the smart phone and requires Apple EarPods with lightning connector to complete (not included). The Sentibo self-fitting software allows the consumer to self-select a hearing profile from a list of options by listening to digitized speech embedded in the software while scrolling through the hearing profiles in real-time through the Apple EarPods. After the setting is selected, the settings can be transferred from the smart phone to the hearing aid directly, though direct audio streaming from the smart phone to the hearing aid is unsupported with the Lumen 155-SF firmware. All Sentibo settings are stored in the application itself, and no internet service is required to perform the self-fitting. The Sentibo application provides 24 predefined options. At the conclusion of the assessment, the software will program hearing aids wirelessly and the user will wear the hearing aids per normal use. The smartphone app will function as a software accessory, allowing the user to make minor adjustments to the hearing aids. The hearing aid is intended to be used with the Sentibo App and to be worn and removed daily by the end user.

Here's a breakdown of the acceptance criteria and the study that proves the device meets them, based on the provided text:

Acceptance Criteria and Reported Device Performance

Study Details

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

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

   • Clinical Performance Study: 93 adults were initially enrolled, with 92 participants providing data for analysis (45 Self-Fit, 47 Pro-Fit).
     • Provenance: Prospective, multicenter, randomized, single-blind, parallel-arm study. The study was conducted adhere to COVID-19 social distancing needs. The text does not explicitly state the country of origin, but the submission is to the U.S. FDA suggesting a U.S. based or internationally recognized standard study.
   • Usability Testing (Human Factors Study): 15 adults (18 years and older) with perceived mild to moderate hearing loss.
     • Provenance: Formative and Summative Evaluations conducted in a one-on-one quiet, comfortable room. No country of origin specified.

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):

   • Clinical Performance Study: The "Pro-fit" arm served as a comparative ground truth, where licensed professionals (audiologists, as implied by the context of "licensed professional-fit" using "commercially available fitting software") fitted the hearing aids. The text doesn't specify the number of individual audiologists or their years of experience but indicates they used "best practice fitting algorithms" and were "licensed."
   • Usability Testing: An "independent, trained moderator" conducted each session. A "cross-functional team" reviewed outcomes, but their direct role in establishing "ground truth" (beyond analyzing usability findings) is not clarified.

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

   • Clinical Performance Study: The study design was a "multicenter, prospective, randomized, single-blind, parallel-arm study comparing outcomes of two fitting approaches: self-fitting... or licensed professional fitting." This implies a direct comparison of two defined groups rather than an adjudication process between multiple expert interpretations of a single case. There is no mention of an adjudication method in the traditional sense of resolving discrepancies in expert opinions.
   • Usability Testing: A "cross-functional team reviewed outcomes at frequent intervals and addressed changes to be considered to labeling, product packaging, and iterative minor design changes to the user interface." This suggests a collaborative review process rather than a formal adjudication method.

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:

   • A traditional MRMC study, as typically applied to image-based diagnostic AI where human readers interpret cases with and without AI assistance, was not explicitly described.
   • However, the clinical performance study does compare "Self-fit" (which uses the Sentibo mobile app for self-fitting) to "Pro-fit" (licensed professional fitting). This is a comparative effectiveness study, but it's not in the context of an "AI assistance" model for human readers. It's comparing two different fitting methodologies for a hearing aid device.
   • Effect Size: The study demonstrated non-inferiority of the self-fit approach to professional fitting.
     • Satisfaction with Sound Quality: Self-fit mean = 2.84 (SD=0.93), Pro-fit mean = 2.72 (SD=1.14) (on a 5-point scale, where higher is better).
     • Satisfaction with Fitting Process: Self-fit mean = 3.22 (SD=0.95), Pro-fit mean = 3.11 (SD=1.07) (on a 5-point scale, where higher is better).
     • SSQ-12 Benefit Score: Self-fit mean = 6.8 (SD=1.7), Pro-fit mean = 6.5 (SD=1.5) (on a 0-10 scale).
     • APHAB Global Score: Self-fit mean = 10.0 (SD=17.3), Pro-fit mean = 12.8 (SD=19.7).
     • The phrasing "similar levels of satisfaction" and "similar levels of perceived benefit" indicates that the self-fit (AI/app-driven) method performed comparably to professional fitting, demonstrating non-inferiority rather than an "improvement with AI assistance" over a human baseline.

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

   • The "Self-fit" arm of the clinical study, which utilized the "Sentibo OTC mobile app" for fitting, represents a standalone algorithm managing the fitting process, with the user's input as part of the algorithm's execution, rather than an audiologist's direct intervention. The user makes selections based on listening to digitized speech through Apple EarPods. Therefore, the "Self-fit" group in the clinical performance study effectively assessed the standalone performance of the algorithm in guiding the user to fit the device without professional assistance.

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

   • Clinical Performance Study: The ground truth for comparative effectiveness was established by the outcomes of licensed professional fitting (Pro-fit group) using "best practice fitting algorithms" and NAL-NL2 targets, against which the self-fit outcomes were compared for non-inferiority. Subjective outcomes (satisfaction, APHAB, SSQ-12) and objective simulated real-ear measures (matching NAL-NL2 targets) were used.
   • Usability Testing: The ground truth was based on the successful completion of tasks by participants and the identification of "critical/essential tasks," reviewed by a cross-functional team.

8. The sample size for the training set:

   • The document does not explicitly state a sample size for a training set for the Sentibo self-fitting algorithm. This type of information (how the algorithm itself was developed or trained on data) is typically for the development stage and not always detailed in an FDA 510(k) summary focused on validation. The Sentibo app offers "24 predefined options" that the user selects from, suggesting a pre-programmed or pre-trained model rather than a user-adaptive learning system that "trains" on user data during the fitting process.

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

   • As the training set size is not provided, the method for establishing its ground truth is also not detailed. However, the description of the self-fitting method for the Lumen 155-SF states it "uses digitized human voice" for selection, and that its fitting method is "adequate for fitting mild to moderate hearing loss, the same as prescribed by NAL-NL2." This suggests that the underlying principles and "ground truth" for the algorithm's output (i.e., the 24 predefined options) are based on established audiological standards like NAL-NL2 or similar validated fitting algorithms for different hearing loss profiles. The predicate device uses a validated NAL-NL2 fitting algorithm, and the subject device's method is compared to this standard.

Ask a specific question about this device