AI and IOT: Redesigning Healthcare in Developing Countries

Ballroom 2

Problem Statement: One of the biggest reasons for the lack of good healthcare in rural areas is the shortage of doctors or skilled nurse practitioners. And most of often than not, people living in rural areas need to travel large distances to visit a doctor. While they only do so for bigger healthcare issues, they cannot afford to do for regular or routine checks

Solution: Follow a proactive or a preventive approach for healthcare in rural areas and that would mean, collecting data about the patient health or medical history, find patterns in the data for the health conditions that they might be susceptible to and then follow a preventive healthcare approach where timely medication and supplies would be available based on the health care issues a particular section of the population in a village might be prone to.

Implementation:

The technological solution or implementation is something that should be easily be handled by a person with limited medical expertise but someone who would have a basic knowledge to handle technology like smart phones and IoT.

Proposed soloution:

  1. Collecting the dataset: We know that dealing with any healthcare data is a very sensitive issue and in real world would run into HIPAA issues. But for the purpose of our demo, we would be using an available dataset in Kaggle.
  2. Training the dataset ( Decision trees) since it would work well with our next integration (which is making it intuitive for end users to use using Dialog Flow). After training, we should essentially have a classifier that would predict what healthcare conditions is the person prone to (based on the boundaries of the collected dataset).
  3. Integrate this to the open source Dialog flow api and perhaps convert it as an action on google home and that would make it easy for any person with a basic understanding of technology to use this to understand the healthcare conditions and perhaps take useful action (like purchasing medications for the commonly found health conditions for the particular village population—taking preventive action)

What can the audience walk away from this?:

  1. For all those who want to solve some pressing problems/issues in the society using their skills as engineers.
  2. Anyone who wants to integrate two different areas of their interest (eg:- someone interested in community service and loves to code or implement new solutions to solve the most pressing societal problems)
  3. Someone trying to understand data science, ML and wanting to get started on it by means of some hands-on project (like me)
  4. Someone who already has expertise in rural healthcare and wants to design public policies on rural healthcare

Technology covered in this session:

  1. Decision trees
  2. Dialogflow api