AI as a potential cure for healthcare’s woes

Todd Soulas
Exploring Delta
Published in
3 min readAug 1, 2017
Photo by jesse orrico on Unsplash

An AI-powered app that listens to your symptoms and offers possible diagnoses to your health problems provides an insight into the future of healthcare. While it won’t be replacing doctors anytime soon, if applied correctly it has the potential to remedy many of healthcare’s woes. Right now healthcare is seemingly at a crisis point in both the developed and developing world — struggling to meet growing demand and at the same time seeking to reduce the cost of services. Of course, AI is no miracle cure. AI cannot address all the issues that exist in healthcare however it can have an impact in several key areas and dramatically improve the situation.

Increasing Access

One of the single most important impacts of AI will be its ability to provide access to healthcare beyond the four walls of a doctor’s surgery or a hospital. With growing penetration of smartphones in developing countries alongside existing mobile phone usage rates, almost universal coverage in developed countries, and the decreasing cost of data, this means that AI-based chatbots and apps have the potential to reach millions of additional patients.

Reducing Barriers

In addition to increasing access to healthcare AI based services can reduce friction for those seeking help and assistance who would otherwise not because many members of the population may not feel comfortable speaking directly with a medical professional (or be seen to speak to a medical professional) at the initial stages. In areas of healthcare that still have an incredible challenge with negative perception, AI based services such as apps may lead to more people seeking help earlier.

Linking to Services

Once a diagnosis is offered, AI-based services can also link you to the right services to seek treatment and further information by suggesting appointment times or even contacting services on your behalf. This alone will have a large impact on improving outcomes in healthcare by decreasing time to treatment and also prompting patients to seek medical advice when they should. This will be particularly relevant for government led awareness campaigns around certain issues such as diabetes, skin checks, and vaccinations.

Better Outcomes

Last year we heard about the use of IBM’s Watson in diagnosing cancer in patients which clearly demonstrated the value that AI can bring to the field of oncology where specialists are under growing pressure to keep up-to-date with dozens of new papers every week as well as dealing with an increasing patient workload (especially in developing countries). While not at the same level, a patient-facing application would be able to draw on a wider range of data over a longer period of time to make recommendations. Linking to included services such as Apple’s HealthKit and other devices that can track patient behavior as well as drawing on previous information provided by the user about symptoms, may lead to better outcomes with diagnosing illness.

What’s next?

The advancement of AI and its application to healthcare has important implications for policy makers and organisations that provide healthcare services (whether public or private).

From a regulatory standpoint this poses many issues — particularly as the technology used to deliver services advances rapidly and easily moves across regulatory boundaries.

Further, similar to autonomous driving technology there is clearly a gap in the environment when it comes to insurance — who is truly liable in the instance of a misdiagnosis if a doctor is using AI as a tool?

Ultimately, despite some issues the role of AI-powered services in healthcare will assist existing medical professionals to do their jobs more effectively and reach a wider number of people — reducing the strain on an overburdened system and increasing standards at the same time. However we will have to carefully manage the development and use of these services to ensure that the benefits to society are maximised and the drawbacks are minimised.

--

--