Starting off in my Physiotherapy career in Sydney 20-ish years ago, I gravitated towards two aspects of the profession – diagnostic reasoning and biomechanical analysis. I also enjoyed combining my clinical work with educational work, lecturing at university and running in-clinic programs.
Fast forward to 2015 and work started on developing a mathematical model. One that could mimic the reasoning process that goes on inside a clinician’s head during the diagnostic process. The complexity of the model was not surprising but the early testing results were quite unexpected. There could be something in this – not a formula to magically spit out a diagnosis based on a single feature or group of features. It was a system that selected relevant questions and was able to “assess” each answer against a greater overall context.
This got around the one dimensional issue with “Dr Google.” A search for “sore knee” brought up the most clicked results. A search for “knee swelling” would do the same. But it provides the most popular per feature. It can’t clarify your answer. It can’t combine all your features into a coherent picture. So the need for a more complex and comprehensive solution is there.
Now to expand the model and see how far it can go…