TECH BYTES
AI HEALTHCARE FROM BIOS
biomarkers, capable of affecting our health. The data is extremely complex, and you need a lot of it to accurately pick out the relevant biomarkers. To speed up this process, Bios created a neural data biomarker discovery platform, which can connect directly into the brain via a neural interface. The system also features a deep learning-based AI system able to identify important biomarkers from the data it receives. In the pre-clinical study, the Bios team recorded the neural data alongside physiological signals generated through more traditional methods such as heart rate and blood pressure. This allowed the firm to secure and synchronise months of continuous neural and physiological data – enough for its AI engine to observe and identify persistent neural biomarkers and their relation to changes in organ function. Oliver Armitage, chief scientific graduate Emil Hewage. Armitage says, “This is the first time we’ve been able to understand the ‘language’ of the nerves as the basis for delivering treatment. We knew we could stimulate a nerve to treat an organ in a similar fashion to conventional medicine, but this gives us the capability to understand and communicate with the nerves and organs directly, so treatments can be made to respond to them in real time.” See bios.health for more information. officer at Bios, co-founded the business with fellow Cambridge
Artificial Intelligence-based treatments could offer a credible alternative to pills when it comes to dealing with chronic conditions such as heart disease. That’s the claim from Cambridge start- up Bios, after it completed a study using AI to detect neural signals that regulate physical biomarkers in the body. Bios, a specialist in neural engineering (or plugging computers into brains) believes the results could open the door for a new generation of AI-based healthcare treatments known as neuroceuticals. By adapting signals from the brain directly, neuroceuticals offer the ability to potentially create targeted and more effective treatments than conventional medicine. Development has been hindered by the speed and accuracy at which scientists could discover and recreate the exact neural signal patterns, or
Graphcore has developed a new type of processor specifically designed for machine learning and artificial intelligence applications. It has already proved popular with customers and high-volume production is ramping up"
43
ISSUE 02
cambridgecatalyst.co.uk
Powered by FlippingBook