EnsoData, Takeda to Improve Narcolepsy Type 1 Diagnosis Using AI Tools and Technology

EnsoData has announced a collaboration with Takeda to jointly develop and validate AI-based polysomnography (PSG) algorithms to support more accurate and scalable diagnosis of narcolepsy type 1 (NT1). Upon successful validation and regulatory clearance, EnsoData will pursue the clinical availability and release of AI algorithms to enable future access, use, and support for healthcare professionals that care for patients with sleep problems. As part of future releases, the algorithms will be integrated into EnsoData’s cloud platform for sleep study data viewing, reporting, and analysis tools to expand AI capabilities and enhance clinical workflows for NT1 diagnosis and follow-up care.

“Today’s clinical and diagnostic tools for sleep disorders, including polysomnography software and sleep study analysis algorithms, have limited dedicated capabilities or support for Narcolepsy or Hypersomnolence Disorders,” said Chris Fernandez, EnsoData’s co-founder and Chief Research Officer.

“These limitations create needs and opportunities for better integration and support for Narcolepsy diagnostic workflows, and AI approaches that could contribute to earlier and more accurate NT1 detection,” added EnsoData’s Chief Medical Officer (CMO), Dr. Nathaniel Watson, Ph.D.

Narcolepsy is a chronic, rare, neurological disorder that results in a range of debilitating symptoms that can include potentially disabling levels of sleepiness during the day. It is currently classified into two different types. NT1 is caused by a loss of orexin-producing neurons in the brain, and narcolepsy type 2, where orexin levels are generally normal.

The NT1 diagnosis journey is long and challenging, often taking up to 15 years for a person to receive an accurate diagnosis. Symptoms often overlap or are confused with more commonly recognized disorders, including sleep apnea, making it difficult to accurately diagnose NT1.

PSGs are an important component of the diagnostic evaluation for NT1, typically capturing 7-8 hours of physiological signals that provide a millisecond-scale view into sleep’s brain waves (EEG), heart rhythms (ECG), respiratory patterns (PPG), and other cardiopulmonary, neurologic, and body movement data. Research has demonstrated the potential to improve the accuracy of detecting and identifying NT1 patterns in PSGs using AI, machine learning, and large waveform databases.

“An estimated forty percent of NT1 patients in the U.S alone who are referred to a sleep specialist and undergo testing remain initially undiagnosed, prolonging an already challenging and complex diagnostic journey,” said Elena Koundourakis, Head of Orexin Franchise Development at Takeda. “Collaborating with EnsoData, we hope to develop and use AI-based tools that could present exciting, new opportunities to alleviate patient burden by improving the diagnostic journey leading to a shorter and accurate NT1 diagnosis.”

Source: EnsoData