What is HOTS?
(Hierarchy of Event-Based Time Surfaces)
Event based predictive algorithm capable of analyzing complex data flows on low bandwidth such as smartphones or tablets, for detection of relevant events & prediction.
HOTS can be ported and integrated on third-party medical devices and mobile platforms for local edge computing and alert generation.
Our core innovation is based on patented machine learning algorithm, which uses a biomimetic approach of computing. Known as HOTS, for “Hierarchy Of event-based Time Surfaces”, the technology has scientifically been validated using computer vision applications. It was recognized to have enormous potential for medical decision support systems because this innovative algorithm can learn to classify complex patterns. HOTS can process any type of incoming signal thanks to its neuromorphic* process. This neuromorphic approach considerably reduces the bandwidth requirement and enables real-time embedded computation leading to a breakthrough of conducting big-data analysis on a regular smartphone.
HOTS benefits and competitive advantages:
Perform data fusion for complex, multiple and continuous data streams
Base its analysis on pattern deviations/global state changes,
Generate alerts to predict any changes in a person’s health status*
Can run on low-power processors and use limited bandwidth.
HOTS can therefore be used for a variety of use cases and applications, on different devices and connected objects (smartphones, tablets, smartwatches and connected devices).
HOW DOES IT WORK
*The clinical trial to demonstrate the safety and efficacy of HOTS to predict Worsening Heart Failure is ongoing with over 500+ patients in Europe.