Telehealth with AI based continuous monitoring and analytics

.

Overview

Thru continuous monitoring and analytics, patients will be informed of potential risks of illness so that preventive measures can be taken.

I will build a dashboard which logs and monitoring incoming data from medical devices. Those data will be inputed to a AI model to do prediction of illness.

The solution will use the IoT dev kit to send data to a back-end where AI model will do prediction using the data. Any illness alerts will be highlighted in the dashboard and patients will be notified of such alerts.

My original design was to use the pulse sensor and DM990013 - CEC1x02 Development Board. But I have problem getting the required debugger/tools for the development with the CEC1702 Development Board. Instead I decided to use the MKR1000 which is based on the SAMW25 Wi-Fi SOC and a temperature sensor for the simulation.

.

.

.

Technology Stack

The following image show the technology stack used and how the components work and exchange data.

.

Setup Grafana and InfluxDB

I am using the stack of Telegraf, InfluxDB and Grafana to make the Dashboard. The Grafana and InfluxDB instance runs in a Docker container. There are lots of tutorial of how to setup this infrastructure. Here is the docker-compose.yml:

version: "2"
services:
  grafana:
    image: grafana/grafana
    container_name: grafana
    restart: always
    ports:
      - 3000:3000
    networks:
      - monitoring
    volumes:
      - grafana-volume:/var/lib/grafana  
  
  influxdb:
    image: influxdb
    container_name: influxdb
    restart: always
    ports:
      - 8086:8086
    networks:
      - monitoring
    volumes:
      - influxdb-volume:/var/lib/influxdb

  networks:
    monitoring:

  volumes:
    grafana-volume:
      external: true
    influxdb-volume:
      external: true

.

Create docker network and volumes:

sudo docker network create monitoring
sudo docker volume create grafana-volume
sudo docker volume create influxdb-volume

.

Create database and user:

sudo docker run --rm \
  -e INFLUXDB_DB=telegraf -e INFLUXDB_ADMIN_ENABLED=true \
  -e INFLUXDB_ADMIN_USER=admin \
  -e INFLUXDB_ADMIN_PASSWORD=supersecretpassword \
  -e INFLUXDB_USER=telegraf -e INFLUXDB_USER_PASSWORD=secretpassword \
  -v influxdb-volume:/var/lib/influxdb \
  influxdb /init-influxdb.sh

.

Use the following command to start the containers:

sudo docker-compose up -d

.

Visit the server_ip:3000 Url to create a database source and build or import a dashboard:

.

.

.

Install Telegraf

Follow this document to install the Telegraf.

.

Setup MQTT

Setup and run a MQTT server in a Docker container:

sudo docker run -it -p 1883:1883 -p 9001:9001 -d eclipse-mosquitto

.

Health Monitor Dashboard

.

Conclusion

The project is partially complete. I just manage to make the ingestion of data to MQTT server work. An AI model yet to be added to do the illness prediction.

.