The system uses existing hospital networks to interconnect camera units in patient rooms to mobile devices carried by nurses to allow hospital personnel to view real-time video of patients. When the system detects a potential bed exit, video-based notifications are sent to nursing staff, allowing them sufficient time to assess the patient and assist with unmet needs that initiated the attempt to exit the bed.
Patient falls are a common, costly, and serious problem in hospitals and care facilities. Conventional fall prevention technologies only alert hospital personnel to a small range of activities that could indicate a broad range of possible situations and patient intentions, and without visual feedback, this decreases effectiveness in both preventing falls and delivering useful and actionable information.
Currently, there is little to no research or data on what patient activity and behaviors most commonly precede falls. In developing our product, we have collected a large amount of data in the form of 3D footage of real patients in rooms during their hospital stays. We continue to analyze this data to gain a better understanding of what precedes and indicates dangerous fall situations.
With improved knowledge of specific patient behaviors and postural changes that are antecedents of unattended bed exits, Ocuvera’s goal is to create an innovative product that gives nursing staff the opportunity to intervene before a potential fall occurs. We have developed preliminary algorithms to detect postural changes (i.e. supine – sidelying – sitting – standing) and predict patient attempts to exit the bed, which are based on many standard and nonstandard computer vision techniques, including: coordinate system conversion, 3D object modeling, motion tracking, background identification, and object segmentation. They also include machine-learned artificial intelligence algorithms such as decision tree risk assessment, decision forest pixel classification, and convolutional neural network classification, all of which are designed by Ocuvera’s computer vision team.
The Automated Video Monitoring approach to fall risk reduction has many advantages. Our predictive algorithms allow our technology to give nursing staff early warning of a potential bed exit, with adequate time to respond. Most existing competitive fall prevention technologies activate only after a fall or bed exit has occurred. Our system also has a low false positive rate and a high rate of accurate bed exit detection. Existing technologies typically have high false alarm rates that lead to nurse fatigue and are thus underutilized for fall prevention. Unlike patient restraints, which are associated with an increased rate of injury without decreasing falls rates, the nature of Ocuvera still allows patients independence and freedom of movement.
Ocuvera also has the benefit of being a cost-effective solution. The mobile nature of our system gives nursing staff the ability to move and set up the intervention quickly and smoothly wherever it’s needed. This flexibility allows hospitals to afford Ocuvera at the level of their specific size and occupancy needs.