But what about health advice, or even medical treatment?

AI and robotics present huge opportunities for the pharma and healthcare industries. According to McKinsey, big data strategies could save the US healthcare system up to $100 billion a year thanks to AI-assisted efficiencies in trials, research, and clinical practice.

So, how is AI being utilised, and will the general public embrace it? Here’s more on the story so far.

Streamlining medical practice

Image classification – which refers to the process of extracting information from multiple images – has so far been one of the main uses of AI and deep learning within healthcare. While the practice does not replace doctors, it essentially makes their jobs much quicker and easier (and lessens the chances of human error).

For example, image classification can help radiologists file and mark low priority X-Rays, taking away the need for them to spend extra time filling in lengthy paperwork.

The technology can aid diagnosis, too. Earlier this year the BBC reported how researchers at an Oxford hospital have developed AI that can diagnose scans for heart disease and lung cancer. One in five cardiologists are currently said to miss problems detected by the timing of the heartbeat. The AI system will be able to pick up on details that doctors tend not to see, also providing recommendations based on the patient’s level of risk.

In the US, AI can also help primary care doctors find and refer patients to specialists – a process that can otherwise be so time-consuming and convoluted that patients often fail to attend eventual appointments. Now, AI systems like Human Dx (Human Diagnosis Project) can create fast, accurate, and actionable insights for doctors and patients alike.

Making preventative care accessible

The general public might remain largely unaware of how AI can aid care in hospitals and doctors’ offices. However, the technology is becoming more mainstream in other ways – mainly through technology that some consumers already own.

Last year, a study found that wearables like the Apple Watch and Fitbit are able to accurately detect serious conditions like hypertension and sleep apnea.

Chatbots are another big opportunity for preventative care, with a number of examples already enabling users to provide advice and even early diagnosis via digital channels. Your.Md and HealthTap are two of the biggest – both using natural language processing to understand common symptoms and help the user come to a conclusion about what might be wrong.

When it comes to health, the big question is whether users will feel comfortable speaking with a bot. Perhaps thanks to our existing reliance on the internet for information (regardless of how questionable) – acceptance is actually quite high.  A report by Reform states that 47% of UK survey respondents would be willing to use an ‘intelligent healthcare’ assistant via a smartphone, tablet, or personal computer, with higher rates amongst younger generations.

With health and exercise-focused apps already proving hugely popular (and users displaying ingrained trust in these brands), perhaps it won’t be long before we see AI being integrated into the likes of Headspace or Nike+ Training Club.

Helping mental health and wellbeing

Another area of opportunity might be in areas that people often find difficult talking about face to face. In this sense, chatbots and digital assistants could potentially even increase accessibility as well as the amount of people seeking help.

Ieso, which is an app that offers cognitive behavioural therapy for managing mental health, is said to have helped over 17,000 people since its launch. Meanwhile, there is also evidence that it has reduced treatment time by 50%. It’s clear that, alongside patients, there are huge benefits for the NHS, with AI assistants helping to take away massive strain on the service.

More recently, Woebot, an app designed to help people cope with feelings of depression and anxiety, was launched on the app store. The idea is that – unlike professional treatment, which can only be sought at a specific time or date – users can interact and access support at any time of day or night.

Alongside preventative care, AI could also help to detect and diagnose mental health conditions. IBM researchers have discovered that machine learning can predict the risk of developing psychosis. By analysing the speech patterns of 59 individuals, it predicted with 89% accuracy which patients would go on to develop a psychotic disorder, as well as detect those who had recently developed psychosis.

Barriers to overcome

While it’s clear that AI is having a big impact, there are still big barriers to overcome before the technology truly transforms the way we access healthcare.

Acceptance of AI to help general wellbeing might be relatively high, however this dramatically falls when it comes to more serious or sensitive issues. Reform suggests that 37% of UK survey respondents say they would use AI to monitor a heart condition, while just 3% say they would use it to monitor pregnancy.

Another area that needs careful consideration is UX. For both doctors and patients, technology needs to be easy to use and intuitive – especially for busy staff who do not have the time or existing skill-set to learn a complicated system.

Lastly, data remains a big issue, with access to good quality data vital in order for the technology to produce accurate results. This comes down to public trust in permitting their data to be used in the first place, as well as granting professionals the right access in the right format.

With the NHS pledging to go paperless by 2020, AI initiatives might be on the back burner while the system struggles with more basic digitisation. However, with benefits that include reducing busy workloads, streamlining processes, and improving accessibility, it’s clear why private pharma companies and start-ups are keen to invest as soon as possible.

Related reading: