Artificial Intelligence, Machine Learning and Healthcare

by Startacus Admin

An excellent example of how AI and ML can be used in health is in the analysis of patient medical records. The points of interest for the AI would be analysing millions of records showing what medication patients are taking, what side effects they experience as well as what illnesses they have.

Here’s where it gets exciting, firstly, the AI would Identify these points of interest using a prediction and recommendation model. For example, with enough patients and their data one could draw conclusions on whether or not the medication was working. In a data set that large, when patients are on the same cocktail of drugs you would be able to identify the negative interactions (which are not currently studied, nor required to be by the FDA). In the same thinking, you could also find positive interactions.

As the data pools became larger, as more people start to use health tracking tools, such as smart watches and home blood test kits the AI analysing it would become smarter; enabling greater predictions and recommendations. Imagine being able to have all of your health data scanned in real-time, along with millions of other people (all anonymously).

You would be able to catch early, and more accurate diagnosis of diseases from Cancer to Parkinson’s and more. As treatments for different illnesses can vary so wildly from Doctor to Doctor, the AL & ML could understand which treatments are most effective, and the reasons for that. As technology becomes more advanced in general and everything we do gets tracked (exercise, diet, so on and so forth) this could also inform how these lifestyle changes affect health and illnesses.

Currently, this is not able to be done, because in clinical trials datasets are far too small, and the level of tracking and data required is far too high to make it possible for humans. AI & ML essentially make it so we’re all continuously in clinical trials; making the treatments we receive more tailored to us as individuals and far more successful. 

Unfortunately the current mantra in healthcare today is to treat symptoms, and not causes, and as such, we throw medicines at the problem; but it may be that lifestyle changes such as diet, exercise and supplements will treat the cause, not the symptoms; eradicating the need for drugs. As much as 50% of people over 50, or even 40, are being advised to take statins. Yet, the number needed to treat, which means the number of individuals who need to take the drug for it to help one person, is 300, “300 people have to take the drug for a year before one heart attack, stroke or other adverse event is prevented”

Some of the ideas talked about here are way off, but they will change the way we think about our health and healthcare. If you have to pay for health insurance, the effect will be reducing its cost; as predicting and preventing health issues will become the norm. In countries such as the UK, it will lift a heavy burden off of institutions such as the NHS, freeing up more time and money, that is so desperately needed. The list goes on and on, and it is going to change the medical industry completely, and indeed every industry. 

About the Author Aaron Brar 

I’m Co-Founder and Director of PLATO Intelligence, a startup which uses AI & ML recommendation and prediction models of text-based data within the recruitment and HR space. I’m personally very interested in AI & ML in healthcare, as well as seeing what our tech can do with it.


If you would like to read more AI related content, you might want to read up about: 

• AI app Lysa which aims to act as your personal health coach offering smart nutritional advice and guidance via your phone. 

• How AI and Chatbots Will Simplify Your Startup Marketing but destroy your customers’ privacy

Subscribe to our newsletter

If you would like to receive our startup themed newsletter, full of the latest startup opportunities, events, news, stories, tips and advice, then sign up here.

• #
• #ai
• #machine learning
• #ml
• #health
• #healthcare
• #aaron brar
• #plato intelligence

Published on: 3rd October 2017