IoT and machine learning-based sensor system for detecting Alzheimer’s Patients movements

Abstract

The deterioration of cognitive abilities is a characteristic of Alzheimer's disease, a neurodegenerative disorder, memory, and behavior Alzheimer’s Disease, eventually leading to loss of independence and eventual death. Despite extensive research, Alzheimer's disease remains incurable, and the available treatments are limited in their efficacy. To resolve the issue, an IoT-based sensor system is proposed that uses microcontroller devices such as Arduino nano, Bluetooth module HC05, pressure sensors, accelerometers and gyroscopes, and machine learning algorithms such as Support Vector Classification (SVC) to check for patients. The system is worn by patients to collect data on their movement and activity. Accelerometer and gyroscopes is used to measure the patient's strength and body, while the Bluetooth module will wirelessly send data to a mobile phone or a central server. By utilizing the machine learning algorithm SVC, the sensor data will be analyzed to ascertain if the individual is walking. This information will be used as an alarm or notification to alert doctors or nurses to unusual situations such as a fall or prolonged inactivity. This information is used by an application for notification or notification sent to the person's mobile. Besides sensor-based monitoring, the system is intertwined with a tailor-made mobile application which informs caregivers about real-time alerts on patient movement hence timely addressing the situation, and minimizing the risks of wandering or falls. Wearable design is focused on comfort and non-invasiveness of the patient, and its continuous usage is focused on safety and dignity. This twofold approach to technical functionality and user-friendliness contribute to the increased practical usefulness of the system in the field of healthcare.