Preventing the Elderly with Diabetes from the Risk of Falling using a Multilayer Perception Classifier

Elderly people are an important age group in society that must be constantly under control. Because of old age, these people suffer from various diseases such as blood sugar, blood pressure, and obesity. If the elderly are not treated and their diseases are not controlled by taking medicine, it will cause headaches, dizziness, and blurred vision, which will cause problems for them as a result of these symptoms. One of the problems they cause is physical and mental injury. Physical injuries in the elderly population are caused by falls. Therefore, providing methods to prevent them from falling can play an important role in their physical and mental health as well as in improving their quality of life. Machine learning algorithms can predict the condition of the elderly at any moment. If an elderly person has acute symptoms, the risk of falling can be prevented by turning on the smart sensors. In this study, we propose a method for predicting falls in elderly individuals with diabetes using a multilayer perception classifier. According to these results, elderly people with diabetes are more prone to falls because of symptoms caused by blood sugar, such as dizziness and motor nerve disorders. By designing an intelligent system that can predict the condition of the elderly based on various diseases, it is possible to prevent them from falling. In this study, various factors collected from this dataset were examined. By reviewing the data and confirming the opinion of a cardiovascular disease specialist, the elderly with diabetes had more symptoms, such as headache, dizziness, and blurred vision, than the rest of the elderly; therefore, these symptoms lead to a drop in blood pressure and fall. The results obtained from the implementation of the machine learning algorithm showed that machine learning could predict the risk of falling in elderly patients with diabetes with a high performance (accuracy=97%).