Women’s health and artificial intelligence

Can Artificial Intelligence (AI) play a critical role in women’s health? Some companies think so.

We all know that women’s health took a backburner to men’s health for centuries and only now is it starting to come to the forefront. About time. And it seems like AI will play a key role. So first, what is Artificial Intelligence? It is an umbrella term for a suite of technologies of which there are essentially four at the top; Machine Learning, Neural Networks, Natural Language Processing and Expert Systems. If you’ve ever talked to Siri, Alexa or Hey Google then that’s Natural Language Processing at work, combined with Neural Networks. If you’re thinking about the Terminator, that’s called General AI and right now it’s impossible and may never be possible.

So then, how is AI being applied to women’s health? In some very interesting ways! One such use is helping prevent Maternal Mortality. The U.S. CDC estimates that 60% of maternal deaths can be prevented. One study in California used AI to analyse medical records and other pregnancy data and they found it could help predict women who are more at risk for complications during pregnancy and into labour.

Another application is using AI to read mammograms, which has been done in the U.S. with success. The doctors help train the AI software along with using millions of previous scans to feed into the system. For AI to work properly it needs a lot of data, like seriously a lot. This could speed up detection and thus getting women into care faster if needed.

The National Institute of Health applied AI to help detect precancerous cells in the cervix as well and found some success. Other uses have been measuring heart activity in women to learn more about how heart attacks occur in women, which we now know can be very different from men.

While this is great progress, there are challenges to overcome. In the past few years it has been determined that there is inherent bias in AI towards women, people of colour and underrepresented groups in society. Largely because AI requires a lot of data and those coding and teaching the AI are often white males. When AI is developed with women and people of colour, the outcomes are much better and far less biased. A recent study by the University of Chicago found that AI in healthcare provided better outcomes and treatment options to white people over black people. Not good.

That these issues have been discovered early on and healthcare companies using AI have taken steps to make changes is encouraging. As we use smartwatches and smartphones to record our physical activities and health data, over time, the use of Artificial Intelligence will only increase and the outcomes will get better.