Pioneering a New Era in Brain Health
Gidi Littwin, a visionary co-inventor behind Apple’s groundbreaking FaceID and Vision Pro technologies, has embarked on an ambitious six-year journey. His mission? To forge a frontier artificial intelligence model capable of deciphering the brain’s intricate electrical symphony, paving the way for early and accurate diagnosis of cognitive disorders.
Littwin’s startup, Hemispheric, has just secured a substantial $52 million in funding, a testament to its progress. The company has amassed an unprecedented dataset from 100,000 individuals, meticulously training deep-learning models to non-invasively examine the brain, eliminating the need for invasive procedures.
From Cupertino to Cognitive Science: A New Calling
Littwin’s departure from Apple in 2020 marked a pivotal shift. His quest for a new challenge led him to a serendipitous LinkedIn message from Hagai Lalazar, Hemispheric’s co-founder. Lalazar, already immersed in developing AI for non-surgical brain analysis, sought a commercially astute partner to propel his vision forward. After interviewing some 75 candidates, he found his match in Littwin.
Littwin’s expertise in large-scale data collection, honed during his time developing FaceID and hand-tracking for the Vision Pro, proved invaluable. “There were massive data collection operations behind these projects, and we knew we had to build something very similar at Hemispheric,” Littwin shared, “and we have.” This foundational experience became the bedrock of Hemispheric’s data-driven approach.
Cracking the Brain’s Code: The Data Revolution
Diagnosing cognitive disorders like depression, Alzheimer’s, and Parkinson’s has historically relied on subjective questionnaires and behavioral observations, largely due to the unique nature of each individual’s brain activity. Hemispheric sought to overcome this limitation by collecting what Littwin and Lalazar proudly call their “most prized possession”: a quarter of a million hours of brain data.
This monumental dataset was gathered from 100,000 paid volunteers across Asia, Tel Aviv, and Boston. Participants engaged in a series of activities designed to resemble games, yet strategically activated different regions of their brains, providing rich, diverse data for the AI’s training.
A Frontier AI Model for Precise Diagnostics
The collected data fueled the development of a sophisticated frontier model. This AI infers brain function from electrical activity within the skull, mirroring how large language models statistically analyze text to deduce meaning. The generalized model has undergone rigorous testing on subsets of individuals, including those diagnosed with PTSD, schizophrenia, and depression, demonstrating accurate deductions about their brain health.
Currently, the team is advancing a clinical study to ascertain the model’s capability to diagnose and even predict Alzheimer’s, promising a future where early intervention could become a reality.
The Future is Non-Invasive: Accessible Brain Health for All
Hemispheric is poised to submit its inaugural product, designed for PTSD analysis, to the FDA for approval early next year, with a public rollout anticipated in late 2027. The diagnostic process is remarkably straightforward: a patient wears a lightweight EEG headset for approximately 15 minutes, interacting with a tablet app to measure brain electrical activity.
Hemispheric’s AI model then assists clinicians in decoding these signals, facilitating accurate diagnoses, guiding treatment selection through predictive analytics, and monitoring patient progress. “The future that we envision is one where this is akin to a blood test,” Lalazar explains. “The device is going to be very, very cheap; it will be able to be sold and distributed throughout mental health clinics, hospitals, and even psychologists’ offices.”
Fueling the Vision: Funding and Strategic Expansion
The recent $52 million funding round, backed by American and Israeli venture capital firms and individual investors like early Uber-backer Howard Morgan, will accelerate Hemispheric’s growth. The capital will be deployed to forge partnerships with governments, healthcare organizations, and pharmaceutical firms, expand its US team, and navigate the regulatory approval process.
In a bid to continually refine its model, Hemispheric plans to gather brain data from millions more individuals. Furthermore, the team is developing its own advanced brain scanners, believing these proprietary devices can yield more pertinent data for their AI models than conventional EEGs. As Littwin notes, “These devices were never built for machine learning and definitely not deep learning.”
As AI giants increasingly venture into healthcare, Hemispheric stands at the forefront, leveraging deep learning to unlock the mysteries of the brain and transform cognitive health diagnostics globally.
For more details, visit our website.
Source: Link
Leave a comment