PrediHealth.ai ; monitoring and predicting disease outbreak

Data could be used to understand the nature of these diseases and how fast they spread. However, only medical related data has been used to solve this problem.  The abundance of alternative big data sources such as transport, social media, climatology allows us to enhance on this already existing method of monitoring diseases to create a robust system that can not only be used to monitor the spread of these diseases but also predict their patterns in real time in order to prevent further spread.

Our solution consists of a mobile app that allows community health workers to collect data and upload data on an epidemic in real time and machine learning model that aggregates this data together with other kinds of data such as transport data to create a robust monitoring and predicting system.

In this challenge, we are working to solve the problem of disease ( especially infectious disease) outbreak.

Infectious disease management is a task held as a high priority by almost all ministries of health all over the world. Outbreaks of diseases pose a huge burden on countries’ government institutions on both economic and demographic fronts thus monitoring and prediction could play a vital role in detection, intervention and preventative measures of such occurrences. Understanding the nature of outbreaks can also help institutions implement health policies to handle future occurrences.

Researchers estimate that  cholera kills 21,000 to 143,000 people annually( source). Ebola killed 11,325 people between 2014 and 2016, 33 people in RDC in 2018 (source ).Tuberculosis as one of the top ten causes of death in the world killed 230,000 children from 1 million children cases in 2017 (source). In Rwanda, tuberculosis has been the main burden wherein 5,755 total cases where notified (source). These diseases listed above are all communicable diseases which are spread directly or indirectly from one person to another.

Our solution has the ability to predict the outbreak of these diseases and to detect them as early as possible.

Our solution is expected to have a big impact on people living in overcrowded locations where the spread of infectious disease would be so fast and very dangerous. It is also expected to have a big and general impact on today’s process of disease monitoring and prediction by collecting more data about disease and using them to make more data-driven decisions. This solution will empower community health workers with the technology to make more advanced decisions.

We have community health workers, health workers, and other colleagues (as citizens)  in the design process of our solution. The major goal for including them was to customize the solution to their needs so that it becomes easy to use and it solves the problem better. We are continuously conducting research with health workers to understand the right information they would like to see on their dashboard, and the information community health workers would access. Our research is not limited to the user experience and user interface research where community health workers, health workers, and other colleagues give continuous feedback on our solution.

Our solution has the ability to connect decision makers, citizens and community health workers with a purpose to monitor and predict diseases. This system has three important parts based on its users:

1. Decision makers:

Decision makers application have a dashboard which communicates with community health workers and citizens applications. The dashboard has the ability to:

• Fetches temporal and spatial data from these applications to make an analysis of an outbreak pattern.

• Send announcements and interact with either citizens or community health workers whenever they need. 

• Predict the disease outbreak in a spatial and temporal way.

• Simulate an outbreak so that decision makers can make an intervention plan whenever there is an outbreak.

• Alert the decision maker when the analysis of the data or the data collected have a potential threat.

This application will receive other data such as transport data, climate data, and other social media data to empower its predictive capability.

2. Citizens

The citizens have an application that has the ability to :

• Locate the nearest community workers and health facilities

• Access information about the health status of a particular region and time.

• Communicate with a community worker or a decision maker whenever necessary

• Access Other health tips 

    
3. Community health workers   

Community health worker have an application that has the ability to:

• Send information about diseases in real time 

• Locate people who need emergent help

• More rich information about the disease to improve their disease diagnosis and about other health information 

• Reporting to decision makers whenever it is possible ( gender-based violence, other types of violence and other incidences)

• Ability to communicate with decision-makers and other citizens

The technologies used :

1. Machine learning

2. Flask, python framework for web application development

3. React Js , Javascript framework for web application development

4. React Native, Javascript framework for mobile application development

5. MognoDB, Database for storing data