About WAVE’s First FDA-Cleared Predictive Algorithm

VSI-screen.jpgVisensia®, The Safety Index (VSI), is The WAVE Clinical Platform’s first FDA-cleared predictive algorithm. Visensia’s Artificial Intelligence (AI) has been developed especially for the advanced and early detection of patient deterioration by our partner, OBS Medical.

Heart Rate, Respiration Rate, Blood Oxygen Saturation, Temperature and Blood Pressure are organized as a correlated set of data, not multiple measures of independent data, providing a more sensitive and specific overview of a patients risk of deterioration.

Visensia® works by using probabilistic modeling techniques to recognize patterns in the data sets that are markers of a patients "novelty" as compared to its machine-learned training data. These markers are able to go far beyond simple  single channel alarms, and crucially can spot when two or more individual parameters look perfectly normal in isolation, but together are abnormal. These markers are almost impossible to manually spot in today’s world of continuous, always connected monitoring for every patient.

With WAVE, every patient gets a graduated 0.0 - 5.0 score based on risk stratifications from comparing patient datasets with the Visensia® “model of normality”.

0 = Normal. 3 = Caregiver attention is necessary. 5 = Critical Crisis.

The VSI is calculated in “real time”, with the vital sign data and their individual contribution to the overall VSI displayed as a pie chart. The data from WAVE can be communicated via your mobility solution/or central station/EMR.

Clinicians can drill-down on any patient to see alarm reports and waveforms for context remotely as if they were at bedside. Patient populations can be score-prioritized by most-to-least critical intervention needed.

AI Powers WAVE’s Predictive Algorithm, The VSI Index

Visensia® is a data driven AI system based on 5 key Vital Signs- HR, BP, RR, SP02, Temp. Visensia® was trained on approximately 3,500 hours of monitoring high risk patients at the John Radcliffe Hospital in Oxford, including patients with Myocardial Infarction, Heart Failure, Respiratory Failure and Elderly post hip fracture. The VSI predictive algorithm has since validated in over 600,000 hours of monitoring.

A VSI Alert occurs either when a single parameter deviates by close to ±3 SDs from the training sets normal value or when two or more parameters depart from normality, but by a smaller amount.

WAVE’s VSI has been shown to have a 95% positive predictive value in identifying patients at risk of Acute Myocardial Infarction, Heart Failure, Respiratory Failure, Elderly post hip fracture and other deteriorations. Read the study*.

While using Visensia in a series of published clinical trial Doctors at the University of Pittsburgh Medical Center (UPMC) went from six unexpected deaths in their control group to zero. Read the study**.

Clinical studies also show that on average Visensia® gives clinical teams 6.3 additional hours in advance of when clinicians would otherwise notice critical patient deterioration.

More Predictive Algorithms in Development

Excel Medical plans to bring other revolutionary predictive algorithms to the WAVE Clinical Platform. Many are currently in development. Please click here if you want to learn more about our partnerships, or to become a partner with Excel Medical

Sources:
* "Integrated monitoring and analysis for early warning of patient deterioration." British Journal of Anaesthesia 97 (1): 64–8 (2006) doi:10.1093/bja/ael113 Advance Access publication May 17, 2006
** "Cardiorespiratory instability before and after implementing an integrated monitoring system."
Article in Critical Care Medicine · October 2010 DOI: 10.1097/CCM.0b013e3181fb7b1c · Source: PubMed