According to CB Insights, healthcare related startups raised a record breaking $4bn in 2019. 

At a time when the world is looking to its healthcare providers to solve the global challenges we’re currently facing, can AI help us reach those solutions faster?

The use of AI as a diagnosis tool has been quite widely covered in the last few months. These solutions often focus on the use of AI and image recognition tech to identify patterns in images, so they tend to relate to diagnoses that are based on the analysis of an image, such as a scan. A good example is Google's recent claim that its AI systems can identify breast cancer in mammograms more accurately than doctors. 

However researchers at MIT have now successfully used an AI algorithm to identify a new antibiotic that's been shown to kill many of the world's problematic disease-causing bacteria - which even includes some strains that are resistant to all previously known antibiotics. 

It's apparently relatively easy to identify chemical substances that kill bacteria. However, it's much more difficult to develop chemical substances that aren't also toxic to our bodies. As a result, all antibiotics introduced in the market in the last 30 years are variations of versions that were discovered in the 1980s. 

Use of AI meant that that the tool was able to screen more than 100 million molecules in a matter of only three days and produce a list of just 23 candidate substances that were: (a) new antibiotics; and (b) predicted not to have any harmful impact on our bodies. That then meant that the researchers were able to reduce a pool of over 100 million molecules to just 23 - which might otherwise have taken years to analyse. 

Whilst there's always a need to keep an eye on safeguards, tools like this could be game changer for the discovery of new antibiotics and drugs more generally. This is a significant move away from AI as a diagnosis tool to AI as a tool for creating cures.

Watch this space...