General Health Check-up, Health Care Model using CNN and Internet-of-Medical Things

Abstract

Many people may believe that undergoing a comprehensive medical check-up is unnecessary when they feel perfectly healthy. However, this perception is mistaken. Regular health check-ups are essential to maintain overall well-being and to detect potential health issues early, enabling proactive steps to prevent them from worsening. Such check-ups are valuable for both individuals in good health and those with existing medical conditions, as they provide insight into the progress of ongoing treatments. Furthermore, routine health assessments aid in identifying risk factors associated with specific diseases. For instance, if elevated blood pressure or cholesterol levels are identified, individuals can receive guidance on adopting healthier lifestyles, such as dietary improvements and regular exercise, to mitigate the risk of heart disease or stroke. The integration of Internet of Things (IoT) technologies has transformed conventional healthcare services and has been pivotal in personalized healthcare and disease prevention initiatives. These advancements rely on the effective extraction of insights from lifestyle factors and activities. Intelligent data retrieval and classification models play a vital role in studying diseases and even forecasting abnormal health conditions. To predict such anomalies, the Convolutional Neural Network (CNN) model is employed, which excels in accurately identifying disease-related knowledge from unstructured medical records. Nonetheless, CNNs can be memory-intensive, especially when utilizing fully connected network structures, and adding more layers can complicate the model's analysis. To address these limitations, we propose a CNN-regular target detection and recognition model founded on the Pearson Correlation Coefficient and regular pattern behavior. Here, "regular" pertains to objects that typically appear in consistent contexts with structurally low variability. Within this framework, we develop a CNN-regular pattern discovery model for data classification.