Saudi Researchers Develop AI Model for Rapid and Accurate Sleep Apnea Diagnosis
In a groundbreaking scientific achievement, Saudi researchers have developed an innovative artificial intelligence model designed to detect obstructive sleep apnea, a condition affecting over one billion people worldwide. This model utilizes unidirectional electrocardiograms and advanced AI techniques.
The study, led by Malak Al-Murshid at the Sleep Medicine and Research University Center of King Saud University, and published in the journal Frontiers in Artificial Intelligence, reveals that the model employs deep learning with attention transformers to provide faster and more accurate diagnoses compared to traditional testing methods.
Results showed that the new model improved the F1 score by 13% over previous studies, achieving the ability to detect apnea episodes with a precision of up to once per second. This capability offers healthcare professionals timely and accurate diagnostic insights while reducing costs compared to comprehensive sleep studies, which can take hours and require manual analysis.
The model uniquely relies on a single biomarker, the electrocardiogram, using intelligent spatial encoding through an autoencoder. This allows it to process raw data without complex prior analysis and ensures efficient operation even in the presence of noise in the data.
This research underscores Saudi Arabia's ambition to invest in artificial intelligence to enhance the medical sector and lead the post-oil knowledge economy, positioning the Kingdom at the forefront of technological innovation both regionally and globally.
