Revolutionizing Brain Health Monitoring with AI

In an innovative leap forward, the Cleveland Clinic has embarked on a groundbreaking partnership with the San Francisco-based startup Piramidal. This collaboration is designed to develop a large-scale AI model aimed at monitoring brain health in patients admitted to intensive care units (ICUs). Unlike traditional systems reliant on text data, this AI will analyze electroencephalogram (EEG) data, obtained through electrodes placed on the scalp that depict the brain’s electrical activity as a series of wavy lines.

EEG readings are invaluable for detecting critical neurological issues, including seizures and altered consciousness. However, current practices in ICUs are constrained; doctors typically generate EEG reports every 12 to 24 hours, making it difficult to monitor each patient in real time. The manual review of a day’s worth of EEG data often takes hours, leading to delays in diagnosing urgent conditions.

Transforming Patient Monitoring

Imad Najm, a neurologist and director at the Cleveland Clinic’s Neurological Institute, points out the limitations of existing EEG monitoring approaches. The process is not only time-consuming but also subjective, dependent heavily on the experience and expertise of the analyzing physician. The system being developed by Cleveland Clinic and Piramidal aims to change that paradigm by quickly interpreting continuous streams of EEG data, flagging abnormalities within seconds. This advancement will empower clinicians to intervene proactively when issues arise.

Kris Pahuja, the Chief Product Officer at Piramidal, elaborates on the system’s intended role: “Our model is designed to monitor patients in the ICU, providing real-time insights into their brain health.” Founded in 2023, Piramidal is on a mission to create a foundation model for the brain that can interpret neural signals across diverse populations.

Supported by Y Combinator, Piramidal has already secured $6 million in seed funding, leveraging both publicly available EEG datasets and proprietary data from the Cleveland Clinic. This multifaceted approach encompasses nearly a million hours of EEG monitoring data from thousands of patients, capturing the significant variability in brain activity. “The foundation model’s beauty lies in its adaptability,” says Dimitris Fotis Sakellariou, Piramidal’s CEO. “It can generalize across individuals, much like how ChatGPT adapts to different writing tones.”

Future Prospects and Implementation

The Cleveland Clinic and Piramidal team is currently using retrospective patient data to fine-tune the model, with plans for an upcoming trial. In the next six to eight months, they aim to test the AI in a tightly controlled ICU environment, gradually rolling it out to encompass the entire unit. This step is expected to allow for the monitoring of hundreds of patients simultaneously, a significant gain in efficiency and responsiveness in critical care settings.

With the introduction of this AI model, the landscape of neurological monitoring in ICUs will undergo a significant transformation. As healthcare continues to integrate advanced technologies, these innovations promise enhanced patient outcomes and a proactive approach to brain health monitoring.