How Will Artificial Intelligence Influence Healthcare’s Next Decade?

Artificial Intelligence is already operating in a range of limited but interesting ways across the healthcare sector. The use of processing computers that can sift and sort data hundreds if not thousands of times quicker than humans is growing, with research suggesting that we spent around $2 billion in venture backed capital on it in 2015. But where is its use likely to impact healthcare in the next decade or so, with reports predicting spending on AI in healthcare will reach as much as  $20 Billion in 10 years time?

Before we look at applications in healthcare in particular, we should remember that AI is an umbrella term for three related technologies; machine learning, extended human cognition and robotics. AI is quite a broad field and in this regard the impact on healthcare as one large industry is likely to be significant, especially in being able to be major new platform/systems leveraging by SAAS systems and databases intelligently talking to each other.

Let’s look at 6 areas within healthcare that seem to have most potential for the use of artificial intelligence.

Clinical Trials

Clinical trials take a long time to process and understand. AI, clever algorithms in particular, can assist with thousands and sometimes millions of datasets. A particular challenge in this sphere is when a doctor are looking to find an appropriate clinical trial for a patient who has exhausted all available treatment options. In this case, artificial intelligence systems can quickly sift through thousands of clinical trial protocols around the entire world, from multiple databases, and determine whether a patient is likely to qualify for the trial and then communicate the results back to the doctor. This fast searching process can readily cover specific patient populations, diseases, conditions, procedures, medications, and much more. 

Diagnosis and Treatment

The volume of healthcare and medical related information is already enormous and is added to and changed every day. In fact, some experts say that medical literature that a physician needs to draw upon to do perform their role well is doubling every two years. This makes it almost impossible for any provider to keep up and only computers can now perform this task in a comprehensive way. A good example of a large computer that can do this is IBM’s “Watson”.  Watson’s AI uses cognitive computing find data patterns that are relevant for a variety of individual patients in order to help them make stronger diagnoses and craft better treatment interventions.


The cultural image of robots is shaped by androids, which look and act like humans. This idea is shaped a little too much by science fiction books and movies perhaps –  in reality the robots of today are much more limited in their scope and focused on very specific tasks where there can be precision on movement. Laser eye surgery was an early adoption of robotic arms, for example, and there are many surgical procedures that can be performed or assisted by robots.  These will not only proliferate but evolve into machines which can give and get instructions verbally and offer medical interventions to patients in a variety of ways. These might be simple at the outset, such as offering educational advice, but will rapidly evolve to highly intelligent response sensitive machines which will be able to do much of what a clinician might do today.

Medication Management

There are many issues in getting patients to adhere to a particular drug or medication regime. These include under-medicating, over-medicating, and drugs or medications going to a person for whom they were not prescribed or intended.

While AI cannot completely eliminate these problems it can help significantly by using several approaches. For example, mobile technology and facial recognition to technology can determine if the right person is taking a given drug at the right time. In addition, patient data from a computer or tablet and draw upon automated algorithms to identify patients, the medication and the process of medication ingestion. Data gathered can be relayed in real-time back through a HIPAA-compliant network where a clinicians can confirm that the patient is taking their medication as directed, or not, as the case may be.

Managing Chronic Disease

When patients have chronic diseases, doctors may use AI to gain better insights to understand their patients’ needs and then tailor care accordingly.

AI can assess disease pathways and why certain people might respond better to certain types of care processes better than others. This is done by not only comparing patient data with others in a database with the same disease but also by taking into account a range of wider data-points such as socioeconomic data and demographics, to help predict how patients will respond to different treatment options. 

Encouraging Wellness

With natural language processing capabilities, AI can assess a patient’s activity goals and readily store these stated goals and then and prompt the person to make sure that he or she meets those goals-with both incentives and sanctions for non-conformance. At one end of the continuum this may be about simple goals around diet or fitness in order to stay well or prevent disease. The other could include predicting which diabetes patients need more help in keeping their blood sugar under control. And it’s not just the providers that can use AI to encourage wellness.

Healthcare payers and Accountable Care Organizations (ACO’s) can also use AI to identify potentially unnecessary services during the review process and improve quality of care. This might include using a database of established clinical guidelines, a specific patient and disease knowledge base, and a predictive analytics algorithm to assign a confidence score that enables consistent clinical and administrative decision-making.

In summary, as we can see, AI is already being used in a variety of ways in healthcare and is set to extend it reach considerably as the technology evolves. Even better, many of these AI solutions (especially the machine learning and cognition parts of it) are now available in the cloud making it both easily accessible and affordable for most providers.

Dr Jon WarnerDr. Jon Warner

CEO-RX4 Group-The Business of Healthcare

Leave a Reply

Your email address will not be published. Required fields are marked *