Anaesthetic risk assessment using a Deep Neural Network which encapsulates the assessments of experienced anaesthetists. Using a rolling time-frame of retraining, ASAi-Risk scores are kept up-to-date while anaesthetic, surgical, and nursing practices evolve. The full methodology is published here.
ASAi-Risk scores can be based on as many or as few inputs as required. The system will warn you if the inputs are insufficient for a score. An ASAi-Risk score goes from 1 (minimal) to 5 (extremely high) in increments of 0.1
ASAi inputs are either discrete or continuous variables: so there can be no dramatic changes to a score when an input value (such as serum albumin) crosses a threshold because such thresholds do not exist.
Previous ASAi networks can be used if you want to calculate a score using the prevailing anaesthetic assessment thinking of an earlier time. Serial test case scores are published here so the evolving behaviour of the ASAi Network can be tracked.