fbpx

How AI Can Predict Arousals in Home Sleep Testing

31.05.2024
Nox Research recently presented a poster at SLEEP 2024 in Houston on the potential benefits of using the deep neural network Nox BodySleep 2.0 for home sleep apnea testing. 

Certain patient populations such as women, non-obese patients, and younger patients are more likely to suffer from obstructive sleep apnea that presents with hypopneas followed by an arousal.¹ These patients’ sleep disordered breathing might not always be picked up in home sleep apnea testing (HSAT), but with increased accuracy through artificial intelligence (AI), more patients could receive a more accurate diagnosis.

Nox Medical, a leader in the field of sleep medicine and AI, recently presented a research poster based on the latest findings at this year’s SLEEP show in Houston on a deep neural network called the Nox BodySleep 2.0.

The research explains how the neural network is designed to enhance HSAT without EEG by predicting sleep stages and arousal events using non-EEG signals. The preliminary findings show that this AI model achieves high agreement with manual scoring for arousal detection by using respiratory inductance plethysmography (RIP), which records changes in the volume of the abdomen and chest.

The researchers believe that by adding arousal scoring to HSAT without EEG, clinicians can improve accuracy in classifying the apnea hypopnea index (AHI) and central AHI severity, reducing the number of inconclusive recordings.

“By improving the accuracy of HSAT without EEG, we can take a step closer towards improving access to care and reducing health disparities,” says Jón Ágústsson, PhD, VP of Artificial Intelligence and Data Research at Nox.

“This is especially important to patients who cannot easily spend a night at a hospital for a PSG study. Furthermore, having repeated HSATs done on the same patient places a burden on the patient and may result in them not pursuing their diagnosis and getting treatment they need,” says Ágústsson.

References: 

  1. Malhotra RK. Arousal-based scoring of obstructive hypopneas. Curr Opin Pulm Med. 2021 Nov 1;27(6):491-495.

Topic: Industry News