Smart Posture Monitoring and Predictive Health Classification for Bedridden Patients Using IoT and AI
Sunitha Cheriyan, Sudha Sakthivel, Thirupathi Regula, K A Varun Kumar, Said Al Riyami · 2024
Posture monitoring has emerged as a critical area of research and application, particularly considering increasing concerns about musculoskeletal disorders and their association with prolonged sitting and poor ergonomic practices. This study investigates the importance of posture monitoring and classification using a comprehensive dataset that includes demographic, and physiological, variables such as age, height, activity level, pain indicators, and personality traits. A pressure mapping system was used to gather data and various AI algorithms were applied to classify in-bed posture using the mined features of the dataset. With different in-bed postures, the study reached high classification accuracy. The performance of these algorithms was then evaluated with machine learning algorithms on the posture dataset. A dataset for posture monitoring and classification study is introduced for accurate and reliable monitoring and classification of these modern health and well-being issues. Relations between features and posture types are identified through the analysis and preprocessing of the dataset. We implemented the classification model using machine learning algorithms and evaluated the model performance evaluation using standard metrics like accuracy, precision, recall, and F1 Score. This paper details current posture assessment methods, providing an overview as well as a background for postural health management and ergonomic solutions. This study aims to propose an implementable posture classification model for a comprehensive assessment of body posture that may help bedridden patients have better health control and a better quality of life. This work might provide more effective care strategies to minimize the risks related to immobility. To achieve, both real-time monitoring and proactive health management, a smart classification monitoring system employs machine learning and Internet of Things technology to automatically identify, classify, and provide actionable feedback on user postures.