Digitalization is nowadays a top topic in popular and scientific debate. In physiotherapy (PT), we are participating in digitalisation on many levels, for example, applications for supporting physically active lifestyle and different systems for distance rehabilitation. Still, there is a lot of potential for development for PTs in this area.
The Artificial intelligence, AI is quite broadly implemented in health care, for example, systems to help making faster diagnosis and giving basic medical feedback. AI can be designed and trained for a specific task, like Siri in an iPhone or AI can be designed for more general purposes covering different cognitive abilities that humans have, and thus, finding its own solutions for an unfamiliar task [Citation1].
Four types of AI have been defined [Citation1]; Reactive machines, Limited memory, Theory of mind and Self-awareness. The Reactive machines (programmes) are those that have no memory, but they can make predictions, that is, what will a consequence be when making a choice in some task. The Limited memory systems have some memory, and thus use experiences for future decisions, for example a self-driving vehicle. The Theory of mind AI-system understands the existence of other’s beliefs and intentions, which then can influence the AI system’s decisions. The Self-awareness AI-system have consciousness of itself implying of understanding how to use AI’s own information in inferring with others’ feelings. The last two types of AI, Theory of mind and Self-awareness have not yet been developed.
We physiotherapists, should be participating in development of these two not so far existing AI systems to improve health care. Human behaviour consists of physical, cognitive and emotional parts. PTs can easily identify the problems in activities in the physical part of behaviour, overt behaviour, by observing and testing an individual. The cognitive and emotional parts of behaviour, covert behaviours, are however harder to understand and interpret. Also, in PT, we struggle with trying to support our clients in adherence to different exercise regimens that in many cases are meant to be lifelong activities, and we do not know enough of how individuals’ beliefs and emotions effect the concept of adherence. To bear these in mind, it could be valuable to be part of teams that are developing Theory of mind and Self-awareness AI-systems. We could increase our understanding of human behaviour and thus be for example better coaches for healthier lifestyle and for supporting behaviour change.
We should also be participating in development of AI systems to meet the demands from future PT-students. Young people are more and more digitalised in their lives as are the future patients and clients. AI is going to influence our lives in the future in great extend. If we educators do not meet the expectations of these students’, they will not seek to our education programmes. The AI could change the students’ way of learning, when, how and with whom they learn. Also, when talking about preparing students to be future health care professions whose daily work consists of maybe fully digitalised hospitals and other care units, we need to match the content of the education programmes with this future image as well. Old school is out, and the new school is in!
There are of course AI-related ethical questions. There is a human bias in training the AI programme, there are hackers in trying to interfere, but also the possible problems coming up by AI finding its own problem solutions. However, if we are not participating in development of new AI systems, we cannot influence the ethical discussions either.
I wish you all a happy AI-year 2019!
Editor-in-Chief
Reference
- Margaret Rouse. AI (artificial intelligence) definition in Essential Guide: Special report: Artificial intelligence apps come of age. Contributors: Ed Burns and NIcole Laskowski. Available from: https://searchenterpriseai.techtarget.com/definition/AI-Artificial-Intelligence (2019 Jan 4).