Cloud algorithm-driven oximetry-based diagnosis of obstructive sleep apnoea in symptomatic habitually snoring...

Area of Science:

• Pediatric Pulmonology
• Biomedical Engineering
• Sleep Medicine

Background:

• Obstructive sleep apnoea syndrome (OSAS) diagnosis in children often relies on polysomnography, which has limited accessibility and high costs.
• Habitual snoring in children can be indicative of underlying OSAS, necessitating reliable diagnostic methods.

Purpose of the Study:

• To evaluate the diagnostic capability of a cloud-driven Bluetooth oximetry algorithm for OSAS in children.
• To compare the accuracy of the oximetry-based algorithm (AHI_OXI) against polysomnography (AHI_PSG).

Main Methods:

• Simultaneous overnight polysomnography and Bluetooth oximetry were conducted in 432 children with suspected OSAS.
• A validated algorithm processed oximetry data to estimate the apnoea/hypopnoea index (AHI_OXI).
• Diagnostic accuracy, sensitivity, and specificity were analyzed across different OSAS severity levels.

Main Results:

• The oximetry algorithm achieved >79% accuracy for all OSAS severity levels.
• High specificity (92.7%) was observed for severe OSAS (AHI >10 events·h⁻¹).
• False-negative rates were low (4.7%), with only 0.6% of cases showing moderate to severe OSAS.

Conclusions:

• Cloud-driven Bluetooth oximetry with a machine learning algorithm reliably diagnoses pediatric OSAS.
• This approach offers a scalable, cost-effective solution, potentially reducing pediatric sleep laboratory access challenges.