Applying Deep Learning for the Classification of Sleep Disorders

An automated classification solutions are essential for treating sleep disorders. Deep learning supports two distinct methods to classify sleep disorders including form-based assessment and real-time video processing. The form-based input analyzes pre-defined sleep-related parameters by using a Decision Tree classifier. The system analyzes consecutive facial data fed through live video with the assistance of OpenCV, NumPy, and SciPy module processing. A Convolutional Neural Network operating within Dlib technology uses the combination of Mouth Aspect Ratio and Eye Aspect Ratio to identify patterns related to sleep disorders. The scoring system generates sleepiness evaluations to classify patients as Sleep Apnea cases or Insomniac or leave them without a diagnosis. This system uses machine learning and real-time data processing to enhance both detection of conditions early on and patient treatment possibilities. The system establishes an automated method for sleep disorder evaluation which facilitates prompt medical assistance while enhancing treatment quality and patient health results.