Present anesthesia machines have many displays and alarms but in fact do little to support the anesthesiologist's decision making in a crisis. The objective of this proposal is to develop an anesthesia workstation which includes an expert alarm system and central display in an effort to significantly reduce the anesthesiologists reaction time to of the anesthesiologist critical events. This system would detect critical events very quickly and provide a clearly defined cause for the problem. Phase I studies demonstrated in animals that an expert alarm system can detect 94% of the anesthesia machine failures and 89.4% of the breathing circuit failures. Phase II studies are planned to complete the development of the alarm system, focusing effort on training the alarms algorithm and developing simple disposable sensors. The alarm system will use neural network type artificial intelligence to identify failures in the patient breathing circuit and in the anesthesia machine. Input to the system will come from a small CO2/flow/pressure transducer array. Phase II includes the animal and clinical testing required to obtain FDA marketing approval.
