In healthcare, there is the added dimension of the application of different technologies, too. These largely revolve around care settings. There are 4 key care settings: Acute/Hospital, Primary/GP, Community & Mental Health, Social Care & Long Term Care (this can be in patient’s own homes, in residential homes or in a number of other places such as day centres or supported housing). This is important as any technology that is applied in a different care setting can lead to very different solutions.
Let us now turn to examine some of the different applications of AI in healthcare and some of the problems they solve. While there are many applications of AI in healthcare, three that have become prevalent in the last few years are diagnostics, triage and self-care (Spatharou, Angela, et al. 2020).
Diagnostics & Triage
The biggest application area of AI in diagnostics is in diagnostic imaging. This means that instead of a clinician needing to check through every image they record for abnormalities, whether it be in radiology, oncology or cardiology, AI technology can analyse these images in the background and identify “normal” and “non-normal” scans very quickly, with a high degree of sensitivity and specificity. This has led to dramatic effects on the workflow of clinicians, by primarily reducing the workload volume and allowing for patients to be prioritised, too.
is an example of a UK company that does this. They analyse X-ray images of the chest and lungs and can provide heatmaps of problematic areas and diagnoses. They have seen growth both in the UK and globally, and are being used to help diagnose and triage COVID patients, highlighting how diagnostic tools are closely related to triaging solutions too.
AI can also be used to diagnose and triage patients before care is delivered. Products known as “symptom checkers” have seen large growth, especially throughout COVID - the biggest names in this space are Babylon Health
. Both companies use AI to analyse a patient's symptoms and characteristics and provide a health diagnosis and then direct them to the best care pathway. Effective use of this tool by primary care clinics means that less patients will be booking appointments, as many cases can be diagnosed and self treated, reducing their workload.
Self care is a growing area of digital health, as it is another means to treat patients outside the practice. Where the application of AI in this area becomes powerful, is when it is combined with remote patient monitoring - being able to track the various different conditions of patients when they are outside the clinic using a wearable device, for example heart rate, blood pressure and more.
Taking the example of Current Health
, a self care platform that uses wearables, the AI in their platform enables changing conditions to be analysed. If a patient shows signs of deteriorating health, the clinician can be alerted before the condition becomes more severe or leads to complications.
What can we expect to change in dentistry?
One of the key takeaways from healthcare is that AI is an extremely flexible tool that has been applied in various care settings to deliver on a wide range of solutions - be it clinician workflow, patient outcomes, early diagnosis and prevention, and more. Nothing is off limits in terms of what the technology may be able to solve - it is simply a case of identifying the problem and finding the right application to use AI to solve it.
As was mentioned in Part 1, Flynotes is already using a strand of AI - Natural Language Processing (NLP) - to reduce the time needed to deliver, and the quality of, informed consent. While many can speculate on how some technology will be applied in dentistry, there is much to be learnt from the healthcare market when it comes to use of AI technologies. It is only a matter of time before we see the rise of various different applications of AI throughout dental practices to solve a wide range of problems.
What problems do you think AI could solve in your dental practice? Leave a comment below!
Mace, S. and Mayer, T. (2008). Chapter 155 - Triage. In: J. Baren, S. Rothrock, J. Brennan and L. Brown, eds., Pediatric Emergency Medicine. Saunders, pp.1087–1096.