AI AND IOT HEALTH CARE MONITORING SYSTEM

Abstract

Continuous monitoring of vital signs is essential for patients with chronic illnesses, elderly people living alone, and patients recovering at home after discharge from hospital. In most cases these patients are checked only during occasional visits to a clinic, and a sudden deterioration may go unnoticed until it becomes serious. Hospitals also face a shortage of nursing staff, which makes frequent manual checks on every patient difficult. This paper presents an AI and IoT health care monitoring system that measures vital signs continuously, sends them to a cloud platform, and uses machine learning to identify patients whose condition is becoming abnormal. The patient unit is built around an ESP32 microcontroller connected to a MAX30102 sensor for heart rate and blood oxygen saturation, a DS18B20 sensor for body temperature, an AD8232 module for a single-lead electrocardiogram, and an accelerometer for fall detection. A small display on the unit shows the current readings to the patient, and a buzzer gives a local warning when a reading goes beyond safe limits. The unit is powered by a rechargeable battery so that it can be worn at home or in a hospital ward without restricting movement. Readings are sent over Wi-Fi to a cloud server using the MQTT protocol at regular intervals. The server stores the data in a time series database and applies two levels of analysis. The first level checks each reading against clinical thresholds for immediate danger. The second level uses a trained machine learning model that considers several vital signs together, along with recent trends, to estimate a risk score and to identify early signs of deterioration that simple thresholds would miss. An anomaly detector also looks for irregular heart rhythm patterns in the ECG signal. Doctors and caregivers access the information through a web dashboard and a mobile application. The dashboard lists all monitored patients sorted by risk score and shows graphs of their recent readings. When a critical condition or fall is detected, the system sends push notifications and SMS alerts to the assigned doctor and family members, including the patient's latest readings and location. Patients can also press a help button on the unit to request assistance.