AI Metric Outperforms Gold Standard for Sleep Apnea Diagnosis

NEW YORK, NY – January 28, 2026 – A groundbreaking study published in the peer-reviewed journal Nature Communications Medicine is challenging the decades-old method for diagnosing obstructive sleep apnea (OSA), a condition affecting nearly a billion people worldwide. Research from medical technology firm Sunrise Group demonstrates that its AI-powered metric, which measures respiratory effort, correlates more strongly with patient-reported symptoms like fatigue and daytime sleepiness than the traditional apnea-hypopnea index (AHI).

The findings could represent a pivotal shift in sleep medicine, moving away from simply counting breathing interruptions and toward a more nuanced understanding of the physiological burden the disease places on patients. This is particularly significant for populations often overlooked by current standards, including women and those with seemingly “mild” cases.

The Limits of a Decades-Old Standard

For years, the diagnosis and severity of OSA have been anchored by the AHI, a score representing the average number of apneas (breathing pauses) and hypopneas (shallow breaths) per hour of sleep. This metric, typically measured during an overnight sleep study known as a polysomnography (PSG), has been the clinical gold standard. However, sleep specialists have long recognized its significant limitations.

A primary issue is the well-documented disconnect between a patient's AHI score and their subjective experience. Many individuals with a high AHI report few symptoms, while others with a low, or even “normal,” AHI suffer from debilitating fatigue and cognitive fog. This symptom-AHI mismatch complicates clinical decisions, leaving many symptomatic patients without a diagnosis or treatment plan.

Furthermore, the AHI is a blunt instrument. It does not account for the duration or severity of respiratory events, nor the degree of oxygen desaturation associated with them. It also struggles to capture the continuous, exhausting effort a person exerts to breathe against a partially obstructed airway throughout the night, even between discrete apnea events. This has led to a growing consensus within the medical community that a more comprehensive approach is needed, with organizations like the American Academy of Sleep Medicine (AASM) actively exploring frameworks that move beyond AHI alone.

A New AI-Powered View of Respiratory Effort

Sunrise Group's innovation aims to fill this diagnostic gap. The company developed a lightweight, chin-worn sensor for at-home sleep studies that analyzes mandibular (jaw) movements using artificial intelligence. From these signals, it derives a proprietary metric called REMOV (Respiratory Effort Measurement from Oximetry and Mandibular movements), which quantifies the percentage of sleep time spent with increased respiratory effort.

The new study evaluated 1,000 adults referred for suspected OSA. Each participant underwent a traditional PSG while simultaneously using the Sunrise device. The results were striking: patient-reported symptoms showed a significantly stronger correlation with REMOV than with AHI or other standard PSG metrics. The link was particularly pronounced among women, a group known to present with different OSA symptoms than men, often leading to underdiagnosis.

“Fatigue and sleepiness matter most to patients, yet symptom severity isn't always reflected by high AHI values,” said Laurent Martinot, co-founder and CEO of Sunrise Group, in a statement. “REMOV offers an accurate view of respiratory effort, helping clinicians understand what’s driving symptoms and identify patients who may be overlooked by AHI-centric assessment. By capturing what AHI can miss, we also realize the true burden of OSA is likely bigger than we thought.”

Critically, the technology has been validated against esophageal manometry, the invasive but highly accurate reference standard for measuring respiratory effort, lending significant scientific credibility to the non-invasive, AI-driven approach.

Addressing the Unseen Burden and Diagnostic Equity

The implications of a more symptom-attuned metric are profound, particularly for the estimated 80% of individuals with OSA who remain undiagnosed. For countless people, the journey to a diagnosis is fraught with frustration. They report classic symptoms of sleep disruption only to be told their AHI is too low to warrant concern or treatment, leaving them to struggle with the consequences of untreated sleep disorders on their health, work, and personal lives.

REMOV offers a path toward diagnostic equity. By quantifying the physiological strain of breathing, it validates the experience of patients whose suffering is currently invisible to standard tests. This is especially vital for women, who may present more with fatigue, insomnia, and depression rather than the loud snoring and witnessed apneas more commonly associated with men. The study found REMOV effectively captures this clinical burden in low-AHI patients, where the symptom mismatch is most frequent.

Adopting metrics that better align with patient experience could empower clinicians to treat the person, not just the number. It promises to bring millions of people out of the diagnostic shadows, providing them with the validation and access to care they need to improve their quality of life.

Shaking Up a Multi-Billion Dollar Market

This technological advance arrives as the global sleep diagnostics market, projected to exceed $1.1 billion by 2032, undergoes a major transformation. The industry is rapidly shifting toward Home Sleep Apnea Testing (HSAT) for its convenience and cost-effectiveness, while AI and machine learning are emerging as powerful disruptive forces.

Major players like ResMed and Philips are heavily invested in integrated digital health platforms and home testing solutions. However, Sunrise Group’s focus on a novel, symptom-correlated metric provides a distinct competitive advantage. Instead of just using AI to replicate the counting of AHI events, the company is using it to generate a fundamentally different and potentially more clinically relevant insight.

This innovation is a core component of the company’s broader strategy, which includes Dreem Health, its U.S.-based virtual sleep clinic available in all 50 states. By integrating its advanced diagnostic sensor into a comprehensive telehealth platform, Sunrise is building an end-to-end ecosystem that can guide patients from initial screening and diagnosis to ongoing treatment and management, a model that aligns with the modern demands of value-based care.

From Peer Review to Clinical Practice: The Path Forward

Publication in a prestigious journal like Nature Communications Medicine is a crucial step in establishing the scientific validity of REMOV, but the journey to widespread clinical adoption requires clearing additional hurdles. While the Sunrise device itself is already FDA-cleared and CE-marked, achieving broad acceptance for a new diagnostic metric involves influencing clinical guidelines and, most importantly, securing reimbursement from payers.

Insurers, including the Centers for Medicare & Medicaid Services (CMS), will require robust evidence demonstrating that using REMOV leads to improved patient outcomes and is a cost-effective alternative or supplement to AHI. The current study provides a powerful foundation for building that case. Sunrise's existing relationships with major insurers through its Dreem Health clinic may also help smooth the path toward reimbursement.

While the AHI is unlikely to be replaced overnight, the mounting evidence highlights its inadequacy. Innovations like REMOV signal a clear and necessary evolution in sleep medicine—a move toward more personalized, patient-centric diagnostics that capture the true, lived experience of a disease that has remained in the shadows for far too long.