This article gives an overview of artificial intelligence, machine learning and deep learning and discusses the potential of artificial intelligence in changing the landscape of healthcare.
Some time ago we wrote a blog post on how so many different factors affect the development of mental health problems in cancer patients. In fact there are so many (intertwined) factors that it is very difficult to identify those who are at highest risk in order to prevent those problems. It is very difficult using traditional methodology because imagine taking into account hundreds of different factors that are interacting with age, treatment, environment in all imaginable ways. At this point, the best we can do is to make mental health services more available for all pediatric cancer patients. But this situation will surely change in the future and this blog post is all about that.
These days we can read almost every day how artificial intelligence (AI) will change the landscape of healthcare – how it can be used to advance diagnostics, to improve treatment, for prevention purposes and so on. There are so many companies working on artificial intelligence to advance healthcare including Google DeepMind Health. Moreover, hundreds of startups are trying to change the landscape of healthcare using artificial intelligence. For example Hungarian startup Turbine is currently creating a solution that has the potential of taking the speed and efficiency of cancer treatment and drug development to another level. Their solution is essentially able to run AI guided simulations to identify complex biomarkers and design combination therapies at an incredible speed. It is a completely personalized approach that wouldn’t have been possible just some years ago.
Intelligent software should be the aim of any healthcare company that would like to stay competitive in following years. Even if you don’t incorporate AI immediately, you should be ready and take it into account from the beginning because data has to be modeled in an AI way. Good tip from experts is to duplicate the data as memory costs are small but HR costs are high. If you want to incorporate AI at one point then be prepared from the beginning. It is definitely the way to go. By the way, healthcare is also the top area of investment in AI as measured by venture capital deal flow.
Artificial intelligence has really come a long way since the 80s when it all started from expert system that was not automated and can be essentially seen as a knowledge base of facts and rules. Now probably the most famous AI is IBM Watson with a knowledge base of pretty much the whole internet to be used for multiple purposes.
To achieve artificial intelligence, machine learning is used. Generally speaking, it is much like a human brain, a network of neurons. These neurons are a blank slate in the beginning. In that situation all neurons start the same and become specialized as they are trained, eventually resulting in specialist neurons that are able to recognize particular patterns of data, able to do more and more complex tasks. Adding more neurons to a network makes it more complex and increases its expressive power.
Deep learning takes a step forward from machine learning. It can train itself with the new data it receives. Many early AI experts were against a situation where a machine can learn from its mistakes just like humans can but this is definitely a reality today. This kind of unsupervised learning is definitely among the recent trends in AI and today autonomous, self-teaching systems are a reality that is revolutionizing many industries.
New ways to optimize machine learning are also constantly developed to reduce the amount of computation. Advances in this field are incredibly rapid.
This topic is very fascinating and it is interesting to see how innovative ways are used to take full advantage of AI. We believe that our solution also benefits from machine learning in achieving fully personalized support for every patient.
It has been estimated that loneliness is the biggest problem in the near future. It affects everyone, not only those who are in the hospital or bed-bound at home. Fortunately people are very eagerly using messenger apps which are helpful in this situation. Actually we have now reached a situation where for the first time ever people are using messenger apps more than they are using social networks and chatbots are taking over the world. Our solution will not only involve our disease-monster-fighting sidekick chatbot and moderated text analysis but also user behavior will be analysed and so on. We try to take full advantage of existing machine learning algorithms in order to advance psychological monitoring and support.
Sorry for a long post! There are so many things to discuss regarding AI and machine learning. We are not experts in AI but we are very interested in the field and eager to incorporate machine learning to our solution. If you feel passionate about this topic then we are always happy to discuss, just contact us!