The Cleveland Clinic and start-up Piramidal are developing an AI model that can interpret EEG data and flag abnormalities in seconds, helping doctors to intervene sooner when monitoring brain health in intensive care units.
The system has been designed to adapt to each individual user, in the same way, ChatGPT can adapt to different types of text.
Following fine-tuning using retrospective patient data, the model will be tested in a tightly controlled ICU environment with live patient data from six to eight months from now.
Piramidal’s brain foundation model has applications beyond the ICU in areas such as sleep and epilepsy monitoring, and brain-computer interfaces.
Questions are already being asked about how and when such brain data should be used, and what ethical frameworks should be put in place to support responsible development and usage.
