Saudi Innovation Reveals Sleep Apnea with Unprecedented Accuracy and Speed
In a pioneering scientific step, Saudi researchers have succeeded in developing an innovative artificial intelligence model to detect obstructive sleep apnea, which affects over a billion people worldwide, using unidirectional electrocardiography and advanced artificial intelligence techniques.
The study, conducted by Malak Al-Murshid at the University Sleep Medicine and Research Center at King Saud University, and published in the journal Frontiers in Artificial Intelligence, indicated that the model relies on deep learning based on attention transformers to provide faster and more accurate diagnoses compared to traditional tests.
The results confirmed that the new model outperformed previous studies by 13% in F1 score, with the ability to detect apnea events with accuracy reaching every second, providing doctors with precise and rapid diagnostic insights, and reducing costs compared to full sleep studies that take hours and require manual analysis.
The model is characterized by using only one vital sign, which is the electrocardiogram, with intelligent local encoding via the autoencoder, allowing for the processing of raw data without the need for complex prior analysis, and ensuring its efficiency even on data containing noise.
This study reflects Saudi ambition in investing in artificial intelligence to enhance the medical sector and lead the knowledge economy post-oil, positioning the Kingdom at the forefront of technological innovation in the region and the world.
