I’m tempered to write that only a few years ago most of the people I know couldn’t imagine a medical chatbot that could work as a nurse. You know, it’s popular to start an article with flashbacks so you would be able to relate the topic to your experience, evaluate the relevance and so on.
Does it have any sense here? I don’t think so. Let’s be honest, nowadays most of the people can hardly imagine a machine that helps in diagnosing illnesses as well.
For many preceding centuries, technological progress used to show up to people who are not from the field by means of major discoveries. The 21st century seems to set up new rules. It insists on the scope of inventions that take place here and there all over the world globe. And you feel helpless in tracing each of them.
From one point it makes you unimpressionable, from another one – you don’t focus on them anymore.
According to the popular nowadays Moore’s law (by the way Moore is one of the Intel founders), a number of transistors in an integrated circuit doubles each two years. It means exponential growth of computer power, which in turn leads to emerging systems that can manage a large amount of data and learn from it. And Indeed, high-powered computers are one of the most significant advantages of the modern technological era.
Just think for a moment about that AI development was stuck for half a century because ML was impossible without processing a lot of data and learning from it. And to handle the data you needed power. Only in the 21st century, we’ve got it enough to effectively build such systems.
Today it’s not only about worldwide companies but about each of us who leverage “smart” automation. We have software that by managing large data sets can help us to effectively research, analyze, schedule, manage, shop, and now even get medical help. At any case, medicine shouldn’t be an exception. It needs automation and machine intelligence at one of the first instances.
It’s not very clear yet which results we can expect from AI development in observable future but undoubtedly it’s going to raise and entrench itself in each of the fields and sides of human living. Of course, it will take time until the majority of people admit usefulness of smart medical assistants embodied in medical apps. Moreover, it will take issues and problems, solved and answered before skepticism ceases.
Fortunately, along with the exponential growth of machine power also diminishes the time to erase skepticism. We do need it to be able to embrace upcoming changes in each field. In this article, we will look at the advancements that have already delved into the medical field but you may not have heard of them yet.
What do you know about the modern medical assistants?
As for me, I don’t know much about human medical assistants who work in medical institutions but I know about the different ones – pieces of software that serve nursing functions. If you’re imagining now robots that look just like real people and work in hospitals with other humans – it’s a wrong way to start. Instead, recall the last time when you have interacted with chatbots. There is a bunch of interesting examples available on Facebook to get your first impression. You can check them here (by the way there’s a medical one among the list).
In fact, chatbots are pieces of software that can handle a lot of data. Now they are helping people to handle many issues by automating routine tasks. I know it sounds pretty boring but in real it’s quite far from that. Large amounts of data contain patterns you are not capable of grasping until it’s processed by something that has enough power. This how software turns into guide into the world of insights you can’t access in any different way.
The idea of medical chatbot designed as a first assistant you communicate with is pretty logical. Every human his own history of diseases. Every disease has a set of symptoms and appropriate treatment. All this is information that can be handled more effectively that it’s used to. Therefore chatbot receives information, analyzes it and passes to the appropriate specialist. Chatbot immediately helps patient to find a local doctor and set an appointment, providing also advice on possible home treatment (if any).
From another point of view, such medical chatbot permanently being fed by information on the histories of diseases and their treatment can also help doctors. It can sound a bit confusing that software can advise doctors which diagnose to pick and which treatment to prescribe. But why not? Hold on for a moment before getting scared.
Consider that usually, it takes years for a doctor to become mature enough to be really great. If human beings are capable of reading (relatively) a lot it doesn’t mean that we can hold effectively all this information in our heads. Not to get overloaded our brains have short-term and long-term memory.
All the information goes to short-term memory. If you don’t retrieve it in short after its acquisition it won’t reach long-term memory and will just vanish. Now think about the incredible amount of information doctors deal with. Don’t you think now that systems capable of quick and effective processing of the data could be helpful for them?
There are no doubts that computers can only assist in curing patients. They can take up functions that don’t require advanced intelligence, which computers don’t possess yet. But they can be helpful in gathering information, figuring out similar patterns and, therefore, helping to find the solution. One of the brightest examples is Baidu app. However, we are going to review the options later.
How doctor app works
We will look at the existing options of medical apps but right now let’s quickly find out the underlying scheme. In a nutshell, chatbot is software that can either rely on hard-coded logic only or contain advanced elements – machine learning, natural language processing, image and speech recognition, and so on.
Taking into consideration the narrow topic of medical chatbots, it might be obvious that only hard-coded logic can’t make them useful. Therefore, they might gain ML and NLP capabilities to handle expected tasks. Therefore, they would be able to understand human language, analyze from input information, and make effective predictions. This is precisely what the most advanced examples can do nowadays.
Medical chatbot to try
You could have heard about it in terms of Baidu doctor app. In fact, the main product of Baidu is Chinese search engine. However, in 2015 they also launched a medical app to widen horizons of users and extend it to quite a peculiar yet always demanding medical field. Later, they create chatbot Melody and built it into the app.
Chatbot’s functions are all about gathering information from patients, directing it to the right doctors, setting appointments, and helping doctors by ordering information and suggesting treatment. Melody is popular even among those who have never tried it. Because its functionality amazes and makes you think about perspectives.
Moreover, Baidu medical chatbot is self-learning system that is constantly improving as the new information is arriving. This is very important not only for the users but also for science. One of such self-learning systems can once become a new breakthrough in AI development.
This is an AI-powered healthcare app with the chatbot inside. It enables users to check symptoms and turn to the proper doctor. The accuracy of the “symptom checker” is more than 90%, which is very high. The company has received investment from Google and they are ready to go further as the system is going only to get better.
Babylon can leave behind young doctors and nurses by its accuracy, easiness, and immediacy of getting qualified medical help. This is what we usually lack in the conventional medical system.
It’s an AI-based app, developed by UK company. It contains a conversational interface (chatbot) for easy interaction that can guide you through the questions about your symptoms and help to find something in their database (for instance if you know a diagnose and you want to check it).
Your.Md was the first healthcare app that included symptom checker service, which is used by customers to get diagnose. The information it’s based on was carefully gathered from the thousands of doctors and proved its accuracy.
Joy is a chatbot application that consults people on mental health problems. It’s available on Facebook messenger (while the rest ones require downloading an app) that makes it even more accessible. This medical chatbot is designed for people who don’t turn to therapists but need to talk to someone. One of the goals of the creators is to show those people that talking about themselves is easy and shouldn’t be a matter of confusion.
The system is based on machine learning and language processing systems of IBM and Microsoft. It doesn’t propose diagnoses but works as someone you can always talk about how you feel. Joy can be used even as a diary to trace the mood during a certain period of time.
Are medical chatbots and apps helpful? Yes. Are they going to substitute real doctors? No. At least in the nearly observable future. The thing is that such medical assistants can be very useful for both parts – patients’ and doctors’. But they can only serve mediator in the patient-doctor relationship.
Information is a serious tool. When you don’t know what is going on with you and where you can go with it, you feel the worst. You will feel much more confident when at the same very moment as you need you can turn for advice. The help at the distance of one click takes away psychological pressure from many patients. At the same time, doctors can benefit from the opportunity to rely on the tons of essential data, which itself can work as a cue in the process of diagnosing and curing.