Detect potential seizures immediately with Convulsions detection.
Real-time video analysis enables quick identification of convulsions as they happen.
Convulsion Detection continuously monitor babies for abnormal and repetitive movements suggestive of seizures. It also detects visual cues for impending potential seizures, serving as a non-invasive and contactless alarm system for swift clinical intervention.
Our studies show that our model is accurate in detecting 95.9% of seizures. Our Livestream video recording works in parallel to provide better on-site documentation and evidence, reducing the time gap in diagnosis and assessment of seizures.
The AI’s data and results can be conveniently viewed in a real-time dashboard, allowing the healthcare team to easily monitor all of their patients at once.
"If we can't hear the babies, our brand new single room department simply operates with the doors open."
The AI analysis provides an automated and objective method of seizure movements, which can be used to supplement the visual assessment of physicians. This can allow large-scale implementation of diagnosis or exclusion of disorders such as epilepsy.
It can be used separately or in parallel. Our S2S package offers live streaming, so this software can record audio along with visual data. Alternatively, it can be used alone and thus record audio alone.
Recorded videos can be transformed into a quantified analysis of infant movements, producing a “stick figure” model. This stick figure extraction allows the infant’s movements to be analysed without the use of their face.