Optimizing Sleep Apnea Care with a Computational Intelligence-Based Health Recommender System for Exercise and Treatment Recommendations

Abstract

Sleep Apnea is a life-threatening condition that occurs when a patient is asleep. During sleep, the patient has recurring episodes of inability to breathe. This paper aims to present a model that identifies apnea and hypopnea events and prescribes the right exercise and treatment to the patient. For this purpose, the proposed method uses a Computational Intelligence (CI) based Health Recommender System (HRS). This new system reviews the patient’s physiological signs, medical records, and sleep data with the help of data analytics and machine learning techniques. The HRS can accurately pinpoint sleep apnea events and offer each patient-specific treatment and exercise suggestions with the help of CI methods. The system maintains the ability to enhance its diagnosis and treatment capabilities over time through the use of adaptive learning. To design HRS, we have employed a collaborative-filtering approach based on k-Nearest Neighbours (kNN) algorithm. We have obtained an accuracy of 86% and an AUC of 0.83 for the apneac event detection. The collaborative-filtering approach computes cosine similarity among patients based on their demographics (age, sex etc.) and Apnea-Hypopnea-Index (AHI). The adaptive learning feature of this HRS depends on feedbacks and ratings provided by patients for exercises and treatments suggested by the system. Lack of dataset on user ratings on medical appliances and data sparsity are main challenges while implementing such HRS. These limitations give new scope for future research work in this field of HRS. The initial outcomes suggest that the proposed method can help provide patients specific therapies faster, which will lead to better patient results. This novel work is planned to change the way sleep apnea is monitored and its treatment is managed. This novel strategy aimed to help physicians, healthcare service providers, and patients who have Sleep Apnea (SA).