HOW REAL-LIFE DATA CAN IMPROVE REAL-LIFE HEARING

Data collection is nothing new to the hearing industry. Individual audiograms and hearing aid settings have always been recorded, and simple data logging via the hearing aids themselves is not a new thing either. 

But, nowadays, data logging is far more nuanced and advanced. It helps analyze environments and links usage patterns to the sound classes that the end-user spends time in and more. Which helps enrich clinical decisions and rehabilitation.

Why data logging benefits everyone

Data logging is a common way to gather anonymous data on usage. It happens with your phone or PC too, and its purpose is to help software developers fix bugs and make improvements, so the user has a better experience. 

That’s also why Widex collects anonymous data. The data contain no personal information from users or hearing care professionals, and they are only used to evaluate the fitting software or apps. This is great because with the data we get smarter and can improve our design and fix issues, so that, in the end, users and hearing care professionals alike have a better experience with our solutions.   

But where do we apply the data? Widex uses data in the whole development process, from defining a feature to designing it and improving it after it’s been released to market. When a feature is considered for an upgrade, we also collect data about this feature to see how it’s used and identify where changes could be made to create a better user experience. Like when we redesigned the process of how to save a personal program in the WIDEX EVOKE™ app. 

The importance of consent to sharing

Data are useful for developing future hearing aid technologies. But when the data are not completely anonymous, consent must be given to share them.

The European General Data Protection Regulation (GDPR) marks a new age in data privacy regulation. While it is an EU regulation, its reach is felt worldwide, since anyone operating within the EU or interacting with EU citizens is required to follow it. 

GDPR concerns any data that can be identified as belonging to an individual, where that individual can be linked to the data. To comply, we’ve ensured a secure and encrypted data exchange via an additional data consent stage in COMPASS GPS. 

To ensure that data are always secure and protected, Widex maintains the highest standards in security and encryption of data and only gives access to a select few relevant employees. Data are pseudonymized, and withdrawal of consent is possible at any time. One good example of how collective data have been used to give back improvements to users of Widex products is from our machine learning feature SoundSense Learn. 

Data from real-time machine learning help improve the user experience 

SoundSense Learn is a feature in the WIDEX EVOKE™ app that end-users can use to adjust their hearing aids 
in situations where they are not entirely satisfied with the automatic settings in the hearing aid. 

SoundSense Learn uses a machine-learning algorithm that asks end-users to listen to a series of pairwise A-B comparisons of different gain settings. When the settings are found, they can be used in the moment, and can be saved as personal programs for future use in similar environments.

This is where data come into play. SoundSense Learn was developed and operates on data from end-user responses. It also generates data itself in the final settings, usage, situations and intentions associated with the individual SoundSense Learn program. These data are interesting for research and development as well as for the hearing care professional. 

Making it faster to get a personalized listening experience

Over time, the SoundSense Learn data have helped developers fine-tune the machine-learning algorithm to identify the ideal settings for the individual end-user faster and more efficiently. 

In Figure 1, we see the maximum number of comparisons that were made to find the ideal sound settings in a given situation. The comparisons or iterations of the algorithm (x-axis) are plotted against the progress to 1.0 (y-axis), indicating that the result is as close to the user’s listening intention as possible. 

SoundSense Learn 1.1 (red) needed on average 17 comparisons to find the best result. When we started learning from the data we collected, we could introduce SoundSense Learn 1.2 (green), which now needs on average as few as 12 comparisons for the same result.

It’s also faster to reach a personal sound: Within 5 comparisons we can see SoundSense Learn 1.2 reaching around 0.75 convergence. In practical terms, this means most end-users experience improvements in just a few comparisons.

In the spring of 2019, SoundSense Learn 1.3 was released. This version added questions on situation (‘Where are you?’)  and intentions (‘What is your hearing goal?’), which the end-user answers before starting the A-B comparisons. 

Both questions come with a range of pre-defined answer options. From the data we can see that almost 50% of the programs are created at home, while other everyday situations are more evenly distributed. 

As you can see in figure 2, there’s also a lot of variation in the intentions across the different situations, which suggest that there are many situations where personalization of sound is relevant. 

These data help us understand the auditory realities of end-users, but the primary purpose of collecting these data is development. Because the knowledge we gain about preferences for the different situations and intentions is valuable input for continued development.

How hearing can become even more tailored with Real-Life Insights

While Widex can use the data for development, it’s also important to get the hearing care professional into the loop about the personal programs their clients create. 

In the latest version of COMPASS GPS (3.4), information on personal programs will (with end-user consent) be included in the hearing aid log that the hearing care professional can review. This offers hearing care professionals a new understanding of an end-user’s real-life hearing, and acts as input for counselling. The hearing care professional can also spot any trends in settings across personal programs, which they can use for more general hearing aid adjustments. 

Overall, Real-Life Insights aims to enrich and strengthen the relationship between hearing care professionals and their clients – with data-driven insights in a user-friendly and informative display, illustrated in figure 3. 

Real-Life Insights requires explicit consent to sharing data

Widex takes data handling very seriously, which is why we ensure a high level of encryption and secure movement of data.

The Real-Life Insights data that move between the end-user and the hearing care professional are handled by Widex and are only visible to Widex if the end-user has given consent to share the data with us. Specific consent is a must for all data exchanges, and at each exchange, Widex ensures the highest standards of security and data privacy.

Helping hearing aids get better

The use of real-life data in audiology and at Widex has come a long way since data were hand-written. And it has a long future ahead of it. Widex has a vision to use real-life data to benefit end-users and hearing care professionals. For this to happen, trust and respect for data are essential. Anyone who chooses to share data with Widex can rest assured that they’re helping develop new hearing aids with new features that will bring new benefits to hearing aid users in the future.

Care to learn more?

Get more details in the full article in Audiology Practices here.