Meet Mandy – An Intelligent and Interactive Medicare System
Summer of Research project by Lin Ni, University of Auckland, supervised by Dr Jiamou Liu.
We’ve all been there – stuck in a waiting room for hours, only to be told by the doctor to go home and rest for a few days. There are numerous nurses and other healthcare staff whose days are filled dealing with patients waiting to receive advice from a doctor. These staff spend their time collecting patient information, booking calendars, and recommending preliminary treatment.
In her summer of research project Lin Ni identified an issue with the current healthcare system in which patients often wait for long periods of time to see a doctor or specialist primarily due to limited resources.
Not all clinics even use questionnaires or forms to get patient information, and if they do, these forms cannot be personalised for each individual. What this means is that the information they provide may not be specific enough for the symptoms they are experiencing, and so the more relevant questions will have to be asked during their consultation with the doctor.
Lin knew there must be a better way. She believes the process can be expedited by using a robot, rather than human healthcare staff. An artificial intelligence (AI) can be trained to ask the same questions a form might require, but then follow up with more specific questions based on the previous answers. Basically, you can give an AI a list of all known diseases, and all the symptoms related to those diseases, and then train it to ask a series of relevant questions before offering a potential diagnosis. A program like this isn’t going to replace healthcare staff, but it can free up their precious time for more meaningful interactions with patients, and help enable clinics to operate more efficiently.
Obviously, a robot isn’t going to provide a completely accurate diagnosis (not yet, anyway), but it will be able to pass on the information it collects directly to the doctor. The robot isn’t going to tell the patient what’s wrong and prescribe his or her medication. Rather, it is an assistant that can help doctors and clinics operate more effectively.
For example, rather than a crowded waiting room, a patient’s first point of contact could be an app on their smart phone. The app could help provide more precise, patient-specific, information to the doctor when they are seen or, if trained correctly, the AI could even offer basic medical advice. Some patients may find the advice given by the AI is sufficient so that they don’t even need to see the doctor, freeing up more resources and furthering the efficiency of the healthcare system.
Lin and her colleagues created such an AI during a PDH summer research project. The AI takes the form of a ‘chatbot’: a persona called ‘Mandy’ that talks to you through an instant messaging application. When opening the app, Mandy will ask for your name, age and gender before the AI will enquire about how you’re feeling. You can respond to Mandy as you would to a human, with a message like “I have a bad cough”. The AI will pick up on the word ‘cough’ and subsequently respond with a series of relevant questions, such as “Are you producing excess mucus?”, or “Do you have abdominal pain?”. After the AI deems that enough questions have been answered, a report is produced that contains all the information you’ve just provided, as well as a list of potential conditions you may be suffering from.
Of course, the capabilities of an AI created in just a few weeks are quite limited. Mandy only has access to a relatively small database of diseases and symptoms. However, the potential of such an application is immense. Imagine an AI like Mandy having access to the combined global knowledge of human health. Not only would the positive health outcomes for individual patients be greatly increased, but the efficiency of the health system itself would be hugely enhanced.
Lin Ni is among a group of students who took part in the summer of research programme funded by Precision Driven Health. This month we are featuring a blog series examining these projects. While at an elementary stage and considered to be a ‘proof of concept’, these projects offer fresh insights into what the world of healthcare will look like when precision medicine is fully implemented.