)
:format(webp))
Five years ago, I joined the team at Starkey as the Chief Technology Officer and Executive Vice President of Engineering to lead our research and development organizations. I was coming from Intel, which is a pioneer in the computing industry and a leading maker of the world’s most advanced processors. As soon as I arrived at Starkey working alongside owner Bill Austin and President and CEO Brandon Sawalich, we began to architect and build what is now known as Genesis AI—our greatest engineering achievement to date. Five years later, I am so proud and excited to see this groundbreaking hearing care technology arrive in hearing clinics across the country.
:format(webp))
Genesis AI hearing aids are revolutionary because they are built to mimic the sensory processing architecture of the brain. Auditory neuroscience research is clear: we don’t hear only with our ears, but with our brain. The cerebral cortex of the healthy human brain effortlessly processes and recognizes sounds by analyzing trains of neural impulses, suppresses noise, and helps us understand speech. This amazing computation is performed by a dense network of billions of neurons within the auditory cortex and beyond. However, our brain loses this remarkable ability when we live with untreated hearing loss, which results in brain atrophy, or loss of neurons, leading to diminished functions.
In order to address this problem, and help our patients hear better than ever before, Genesis AI features an unprecedented upgrade to the brain of the hearing aid: the Starkey Neuro Processor, our industry’s most advanced processor technology. Just as the basic computing unit in the human brain is the neuron, in processor chips, it’s the transistor. Built on advanced semiconductor technology, the Neuro Processor features 6 times more transistors than previous hearing technology, allowing it to compute 4 times faster, and providing 5 times more processor memory and 10 times more system memory. This processor chip is a once-in-a-decade upgrade. It's not a minor or even incremental improvement from the processor of the past generation devices or even comparable to technology that our industry is currently familiar with.
:format(webp))
In addition to that, we have built a deep neural network engine right into the Neuro Processor. It is a computational hardware block in the chip that is architected like the cerebral cortex of the human brain, processing information with a network of interconnected neurons. This onboard AI engine enhances speech and suppresses noise, like the healthy human auditory cortex does.
When we’re looking at the daily lives of hearing aid wearers, the world throws very complex acoustic environments their way — from having a conversation in a relatively quiet environment, to meeting friends in a noisy cafeteria, to walking out in the street where cars are honking and there's wind blowing — and the device has to learn to deal with it. Previously, engineers had to handcraft specific settings and algorithmic tweaks to make a hearing device behave, for people going about their daily lives. What AI, and specifically deep neural network, does is a data-based approach for the device to become smarter and deal with challenges that are thrown at it from an acoustic viewpoint by learning from the vast amounts of data.
For decades, the focus has been on what is lost in the ear. The neurobiology of the ear is very well known where it essentially performs the functions of a directional microphone. The cochlea performs tonotopic frequency analysis and the outer hair cells provide nonlinear amplification. Finally, you have the transduction process that converts sounds into neural pulses that are sent onto the brain to process the sound. The brain is where the magic happens — understanding conversations, reducing noise, and helping us pay attention to what we care about or should care about. That happens up in the cortex. This amazing computation suffers if we live with hearing loss for a long time. When the ear isn’t sending the brain the same amount of stimulation due to untreated hearing loss, you have brain atrophy which makes it more difficult for the brain to understand sound, even if we make it louder.
:format(webp))
The opportunity for processing that goes on in a hearing aid now is not just to amplify sound in a way to mimic the sensory cells in your ear, but also, to pre-process that signal in a way to help your brain understand the sounds that matter. This is an opportunity for us to perform processing faster, smarter, and more efficiently with an all-new additive compression system to make it easier for our patients to understand speech, reduce noise, and just go about their day-to-day lives.
This is a ground-up redesign for the overall listening experience. Patients are going to notice the difference immediately. They should feel reduced listening effort in noisy environments and an improved ability to understand conversational speech in difficult environments. They’ll be able to enjoy very soft sounds and avoid loud sounds from reaching an uncomfortable level.
In addition to helping people hear better, Genesis AI offers multifunctional benefits that enhance daily life. It can stream audio from smartphones and other devices, track health and activity data, alert designated contacts if the wearer falls, and act as a personal assistant that answers questions and translates languages. On top of all that, Genesis AI delivers an unprecedented 50+ hours of battery life on a single charge.
Genesis AI is a new generation of hearing healthcare technology. All-new everything, future-proofed to build upon for years to come. This is just the beginning.
:format(webp))
To learn more, you can listen to my conversation with Starkey Chief Innovation Officer Dave Fabry, PhD, on a recent episode of Starkey Sound Bites.
Achin Bhowmik, PhD, ist der Chief Technology Officer und Executive Vice President of Engineering bei Starkey Hearing in Minneapolis. Dr. Bhowmik ist verantwortlich für die Technologiestrategie des Unternehmens, globale Forschung, Produktentwicklung, Engineering und Programmmanagement-Organisationen. Zuvor war er VP und General Manager der Perceptual Computing Group bei Intel Corp und ist derzeit auch außerordentlicher Professor an der Stanford School of Medicine.
