As an example of how this technology might be used in future clinical practice, Abiomed has already trained an AI algorithm to predict the next five minutes of a patient’s arterial pressure based only on the prior five minutes of console data. (see Figure 1) Abiomed has also developed AI algorithms to predict other parameters, such as stroke volume, left ventricular pressure and cardiac output. The AI algorithms are not yet cleared or approved for patient use. Once fully developed, they will be submitted for regulatory review.
Predictive analytics are possible by integrating Impella clinical study data with Impella console data from thousands of cases and training artificial intelligence networks on the co-registered data. AI networks could then receive and analyze console data in real-time and send patient-specific predictions to that patient’s medical provider. (see Figure 2)
“Artificial intelligence networks, properly trained using large volumes of streaming data, can be powerful tools to aid in clinical decision-making,” said Chuck Simonton, MD, Abiomed’s chief medical officer. “One day, using artificial intelligence, physicians may be able to confidently predict a patient’s future hemodynamics. That would make clinical decision-making more efficient and improve patient outcomes.”
Abiomed is also studying artificial intelligence to make more holistic predictions, such as the probability a patient will recover his or her native heart function. This information could help medical providers determine if an alternative course of action is needed.