Latest news with #Piramidal
WIRED
6 days ago
- Health
- WIRED
An AI Model for the Brain Is Coming to the ICU
Aug 11, 2025 11:00 AM The Cleveland Clinic and startup Piramidal are developing an AI model trained on brain wave data to monitor intensive care patients. The Cleveland Clinic is partnering with San Francisco-based startup Piramidal to develop a large-scale AI model that will be used to monitor patients' brain health in intensive care units. Instead of being trained on text, the system is based on electroencephalogram (EEG) data, which is collected via electrodes placed on the scalp and then read out by a computer in a series of wavy lines. EEG records the brain's electrical activity—and changes in this activity can indicate a problem. In an ICU setting, doctors scan EEG data looking for evidence of seizures, altered consciousness, or a decline in brain function. Currently, doctors rely on continuous EEG monitoring to detect abnormal brain activity in an ICU patient, but they can't monitor every individual patient in real time. Instead, EEG reports are typically generated every 12 or 24 hours and then analyzed to determine whether a patient is experiencing a neurological issue. It can take two to four hours to manually review a day's worth of brainwave data. 'This type of thing is time-consuming. It is subjective, and it is experience- and expertise-dependent,' says Imad Najm, a neurologist and director of the Epilepsy Center at the Cleveland Clinic's Neurological Institute. The system that the Cleveland Clinic and Piramidal are developing is designed to interpret continuous streams of EEG data and flag abnormalities in seconds so that doctors can intervene sooner. 'Our model plays that role of constantly monitoring patients in the ICU and letting the doctors know what's happening with the patient and how their brain health is evolving in real-time,' says Piramidal's chief product officer Kris Pahuja. Pahuja and CEO Dimitris Fotis Sakellariou founded Piramidal in 2023, with the goal of building a foundation model for the brain—an AI system that can read and interpret neural signals broadly across different people. Prior to this, Sakellariou spent 15 years as a neuroengineer and AI scientist doing EEG research. Pahuja worked on product strategy at Google and Spotify. Their startup, which is backed by Y Combinator, raised $6 million in seed funding last year. The company built its ICU brain model using publicly available EEG datasets, as well as proprietary EEG data from the Cleveland Clinic and other partnerships. Sakellariou says the model incorporates nearly a million hours of EEG monitoring data from 'dozens of thousands' of patients, both neurologically healthy and unhealthy. Brain activity patterns are extremely variable from person to person, so building a brain foundation model requires huge amounts of data to capture common patterns and features. 'The beauty of a foundation model is the same way ChatGPT can generalize text, it can adapt to your tone, it can adapt to your way of writing—our model is able to adapt to the brains of different people,' Sakellariou says. Currently, the Cleveland Clinic and Piramidal team is using retrospective patient data to fine tune the model. In the next six to eight months, they plan to test the model in a tightly controlled ICU environment with live patient data and a limited number of beds and doctors. From there, they aim to slowly roll out the software to the entire ICU. Eventually, the software will allow the hospital system to monitor hundreds of patients at once, Najm says. The slow rollout is to reduce the rate of false positives and false negatives—instances where the system misidentifies patients who don't have a severe event or failing to catch someone who does. The latter scenario especially is 'a big problem that keeps us awake at night,' Najm says. Piramidal did not comment on the model's current accuracy but said it has evaluated its technology against a network of doctors and has achieved 'human-like' performance. The company plans to publish data on the model's accuracy at a future date. While Piramidal's immediate focus is applying its brain foundation model to the ICU, Sakellariou and Pahuja also want to use it for epilepsy and sleep monitoring. Meanwhile, brain-computer interface company Synchron is developing a brain foundation model incorporating data from trial participants to make its system more accurate and generalizable to more users. There are also consumer applications of brain foundation models, such as using EEG earbuds to measure emotional states. Both medical and consumer applications raise questions about how brain data will be used and stored, as well as how and when it should be used. 'Advancements like this one highlight the need for anticipatory ethical frameworks that support responsible development and use of these technologies,' says Caroline Montojo, president and CEO of the Dana Foundation, a private philanthropic organization dedicated to neuroscience research. 'It's critical to bring in many different perspectives at early stages of technology design from multiple disciplines, including ethicists, social scientists, and legal scholars, as well as the lived experiences of patients.'
Forbes
10-08-2025
- Health
- Forbes
Doctors Use Large Neuro Model To Decode Brain Activity
While Dimitris Fotis Sakellariou and Kris Pahuja both shared a passion for playing music, what ultimately brought them together was an opportunity to use artificial intelligence to advance the field of brain science. Sakellariou's medical research and deep technical skills coupled with Pahuja's AI strategy and product credentials were the perfect mix and in 2023 they became cofounders of Piramidal. As a graduate startup of Y Combinator, what has made Piramidal particularly compelling is that they have built a large foundation model that instead of learning from a corpus of text, uses data produced by electrical activity in the human brain. In this way, their AI is trained to understand and detect patterns of brain language potentially transforming neurological diagnostics. It's the first step in many that they hope will lead them to their ultimate goal of building a fully AI-enabled neurologist. Building The First Large Neuro Model In November 2022, ChatGPT was released to the public. It enabled anyone to type a plain English question into a text box and get a natural sounding, informed response. ChatGPT was most people's first encounter with a large language model, a type of AI. In simple terms, an LLM works by being trained on a massive amount of text data that is derived from websites, databases, articles, and more. Through this process, the LLM learns language patterns and is then able to apply them in response to input from a user. Sakellariou, who holds a PhD in neuroscience and AI, had a breakthrough idea to build a specialized LLM, which his team now calls a large neuro model, that would use data, specifically neural language from the brain, from an electroencephalogram also known as an EEG. EEG devices, found in a clinical setting, conduct tests that record and display brainwave patterns, and are used to detect and investigate epilepsy, and other problems such as dementia, brain tumors, sleep disorders, and head injuries. What Problems Can A Large Neuro Model Solve? In a typical hospital context an EEG is hooked up to a patient through electrodes that are placed on the scalp. Brainwaves are displayed on a monitor or printed on paper. Doctors, nurses, and other medical technicians check on EEG results from many patients periodically during the day. As a practical matter, it's not possible for medical staff to continuously monitor and interpret EEG output. As an example, if a doctor checks an EEG in the morning and then before lunch the patient has a brain dysfunction, the doctor may not know about it until they check the EEG again in the evening, when appropriate intervention may be too late. New York-based Piramidal's LNM solves this problem by constantly consuming the EEG data, enabling it to produce accurate patient time series reports, in seconds. The LNM's on-going monitoring means it can analyze, identify, flag, and alert medical staff about abnormalities in real-time. Treatment close to or as the medical event is occurring can literally save the patient's life. Their model also eliminates the manual time-consuming work required to study EEG results, which often takes hours of effort, and it is particularly valuable in situations such as emergencies, when high quality data can support better real-time medical decisions and interventions. The result? Improved healthcare outcomes. Cleveland Clinic Makes A Bet On Piramidal Cleveland Clinic, opened in 1921, is a medical center with 23 hospitals and 280 outpatient facilities globally. In 2024 it served close to 16 million patients, and it is considered one of the world's top centers for neurology. As a large provider, the Clinic has around 100 EEG devices in ICUs serving patients at any time. Monitoring, reporting, and managing each EEG is a highly time-consuming task relying on scarce time availability from medical professionals. In addition, the current absence of real-time brainwave time-series analysis, interpretation, and alerts means inefficiencies can exist in being able to reduce brain injury and even death in the event of an ICU emergency. It makes sense then that Clinic leaders would have a keen interest in Sakellariou and Pahuja's innovation and consequently, a strategic collaboration is now underway. Over a period of several months, Piramidal's LNM will be deployed across many of the Clinic's ICUs. The center will work to co-develop a custom version through testing and refinement that meets their specific needs. Sakellariou believes the solution that emerges from this collaboration will also inform the development of a more widely available commercial version for medical networks across the world. A Challenging But Bright Future For AI In Healthcare AI is ushering in a new era of healthcare innovation. Today, breakthroughs using AI in multiple areas of medicine are happening with greater frequency. Examples include greater accuracy in imaging and diagnosis, acceleration of drug discovery and development, robots assisting with surgery, and precision medicine enabling treatments to be tailored for each patient. There's a lot happening to be encouraged and excited about in the medical field. That said, it will take more than just advances in technology to realize the benefits of innovation in healthcare. Pahuja sees many non-technical hurdles in the way, particularly in the US. Despite the availability of solutions, slow technological adoption is still a characteristic of healthcare systems for many complex reasons including the process in which reimbursements are made. In addition, the current healthcare regulatory environment can quickly become a roadblock for adoption of AI. Despite these hurdles, both Sakellariou and Pahuja are convinced that healthcare innovation driven by AI is about to flourish, and they are well positioned to ride what will likely be a long wave. They acknowledge that it's going to require a lot more investment, after all, training a large neuro model doesn't come cheap. With AI, perhaps many of our worst healthcare fears, from cancer to neurological diseases, will be soon be overcome. That future can't come fast enough.
