Opioid-Overdose-Prediction-System

A app to prevent any event of opioid overdose

This project proposes an innovative and out-of-the-ordinary healthcare system called "Opioid overdose Prediction System" to foretell the risks on one's life put on by their prevailing health predicaments. On top of that, we will work to provide an elaborate data-driven real-time medical-alert wrist-band for the person at risk of overdose. The aforesaid wrist-band will use blood pressure, Sp02, breathing rate sensors among other necessary hardware, and work synergistically with our Machine Learning model to instruct the people in concern, the precise technique, and accurate dosage of the crucial opioid inhibitor - "Naloxone" to be administered to the person in need through our dedicated app. The mobile application in our consolidated system will also keep track of the personal inventory and availability of Naloxone in nearby pharmacies for pre-emptive emergency necessities and prompt action.

Objectives

● To develop a hardware device that analyses various body parameters that cause respiratory depression.
● To check if opioid overdose is the reason of respiratory depression.
● Build an opioid overdose assistance application that will Notify the nearest contact
● Tell the exact amount of naloxone to be given at a particular instance
● Provide details about where to purchase naloxone and the contact of the nearest medical service available.
● To ensure the app is user friendly to ensure that the user faces no difficulty
using it.

Project Outcomes

We will Build a Mobile application that will be able to detect the respiratory depression induced by opioid overdose with the help of a wrist band .

1. Wrist band will be able to take reading of heart rate Spo2 level and ECG readings using sensors like MAX30100 and AD8232 respectively.

2. We will be using ESP8266 wich is a NodeMCU based micro-controller. It works on Espressif ESP8266-12E Wifi System-On-Chip to provide a smooth connection with the database server.

3. We will be using Firebase Database to store our data so it will be safe on a cloud server and also to maintaine a stable connection between our band and mobile application.

4. The mobile application is developed using react native technology so we can make app for both IOS and Android and further we have used a redux based approcah for deployment of application as it is good for scaling and it is commen as per industrial stenderds.

5. We are using a Machine Learning based model to predict the risk persentage for a user by taking input as user medical history and current medication and this will help us to map user's values for a case of opioid overdose.

6. Machine Learning model is hosted on heroku server and we have used it as a api to connect it with the mobile application.

7. In mobile application user will be provide with graph representing his vitals reading obtained after connecting the watch with the application.

8. Mobile application will also provide features to see Naloxone availablity wich will be implemented by google maps api.App has also a feature to show correct procedure to consume a naloxone dose .

9. Mainly whenever our app detects an opioid overdose senerio for the user it will trigger an alert and if user fails to check this alarm as false detection app will notify all user's friend with the current location of the user.

Product Perspective

1. First step on app is to login. User can sign in
   with mobile no or email but they have to give
   mobile number and they also have to give
   required information like –
   ● Current Name and Address
   ● GPS and Bluetooth service allowance
   ● Name and contacts of people to inform

2. On Second Step User have to answer some
   questions that will help us in prediction of his
   opioid overdose risk. Answer to be given
   either Yes/No.
   a. Bipolar disorder or schizophrenia?
   b. Heart failure?
   c. Chronic kidney disease with clinically
   signicant renal impairment?
   d. Stroke (cerebrovascular accident, CVA) or
   other cerebrovascular disease?

3. On third step user have to give details of
   his/her ongoing prescription so we can
   further predict the opioid risk and can analyse
   the user in better way to give our best help.

4. On Step four app is ready to use the user have
   to connect the watch with the app through
   Bluetooth and their user can monitor his ECG
   rate and SpO2 level and we will calculate
   opioid risk at every time.
   Here User is provided with other options
   such as –
   ● Naloxone availability nearby.
   ● Opioid overdose prevention facts and articles.

5. If risk rate goes above a threshold value, we
   will give an alert on mobile and if due to any
   error this alarm is caused the user will be
   given 30 seconds to stop it, else an alert is
   send to all contacts given by user and further
   assistant is given to them to save user’s life
   from overdose.

Hardware

Screenshots