Towards a Wireless Smart Polysomnograph Using Symbolic Fusion.

Area of Science:

• Medical Devices
• Biomedical Engineering
• Sleep Medicine

Background:

• Polysomnography is the gold standard for diagnosing sleep disorders like Sleep Apnea Syndrome (SAS).
• Current polysomnography faces challenges in reliability due to time-consuming manual scoring with high inter-scorer variability and wired sensor artifacts.
• Sleep Apnea Syndrome (SAS) is a public health concern linked to increased cardio- and cerebro-vascular risks.

Purpose of the Study:

• To develop and validate an automated method for Sleep Apnea Syndrome (SAS) diagnosis.
• To assess the performance of symbolic fusion in sleep event scoring based on AASM guidelines.
• To enhance the reliability and efficiency of polysomnography for sleep disorder diagnosis.

Main Methods:

• Utilized symbolic fusion techniques for automated sleep event scoring.
• Applied the international American Academy of Sleep Medicine (AASM) guidelines for scoring sleep events.
• Evaluated the method on a sample of 70 patients, focusing on the Apnea-Hypopnea Index.

Main Results:

• Symbolic fusion demonstrated a high level of agreement (97.1%) with expert sleep scorers for the Apnea-Hypopnea Index.
• The automated method shows comparable performance to human experts in diagnosing Sleep Apnea Syndrome (SAS).
• Preliminary results indicate the potential of symbolic fusion to overcome current polysomnography limitations.

Conclusions:

• Symbolic fusion is a promising approach for the automated diagnosis of Sleep Apnea Syndrome (SAS).
• This method offers a reliable alternative to manual sleep scoring, reducing variability and potential artifacts.
• Future work includes confirming these findings and developing a smart wireless polysomnograph system.