Melbourne-based brain-computer interface startup Fluent has raised AUD $2 million to advance its communication technology for speech-impaired individuals. A spinout from the University of Melbourne, the company is developing a minimally invasive device to translate thoughts into speech. This funding will propel the solution towards human clinical trials, offering hope for people with neurological disorders.
A New Approach to Brain-Computer Interfaces
Individuals with conditions like motor neurone disease lose their ability to speak, often relying on cumbersome communication aids. Current technologies involve slow, manual selection of letters on a screen through touch or eye movements. These methods can be frustrating and limit the natural flow of conversation for many users.
While advanced brain-computer interfaces exist, they typically require invasive and costly surgery to implant electrodes directly into the brain. Fluent is pioneering a lower-risk alternative with a device inserted just under the scalp, but outside the skull. This approach, comparable in safety to a cochlear implant, aims to make the technology more accessible.
Decoding Thoughts into Speech
The device functions by monitoring the motor cortex, the brain region controlling speech muscles. Cofounder Dr. Tim Mahoney explains that when a person attempts to speak, their brain generates unique electrical signals. Fluent's interface is designed to capture these neural "QR codes" in sequence to interpret intended speech.
This captured brain activity is then processed by a sophisticated machine learning model. The model's purpose is to convert the complex neural signals directly into audible speech or written text. This will enable non-verbal individuals to communicate their thoughts seamlessly, without needing to type or push a button.
Validation and Future Steps
The technology has shown significant promise in preliminary human testing at St Vincent’s Hospital Melbourne. In these trials, participants wore a cap with 144 electrodes that recorded brain activity as they spoke or imagined phrases. This research was instrumental in building a large dataset for training the AI model.
By collaborating with a Japanese team and leveraging a larger dataset, the system achieved an impressive 96% accuracy in identifying phrases. This result validates that high-quality signals can be captured without invasive surgery. The company now plans to begin clinical studies of its insertable electrodes later this year.
Strategic Investment and University Support
The $2 million funding round was supported by a diverse group of local and international investors. Backers include the University of Melbourne’s Genesis Pre-Seed Fund, Galileo Ventures, and Multiple Sclerosis WA. International support came from New York’s Jumpspace Ventures, London’s Founder’s Factory, and Auckland's Pacific Channel.
Professor Mark Cassidy of the University of Melbourne emphasized the startup's origins within the university's innovation ecosystem. He noted that Fluent was nurtured through its accelerator program and received crucial proof-of-concept funding. This support was vital in preparing the company for human clinical trials and commercial development.
This successful funding round marks a pivotal moment for Fluent, providing the capital to advance its life-changing technology. The investment will accelerate the transition from promising research to tangible clinical applications. Ultimately, Fluent's minimally invasive brain-computer interface holds the potential to restore communication and independence for countless individuals worldwide.