Artificial Intelligence in Healthcare: What it can do for us?
17 October 2018
Blog by Elaine Lima da Souza
We are living in the Age of Data. Everything is happening fast, and we are not able to keep following up with all information available. We need to be assisted by technology. Artificial Intelligence (AI) can help us to make a good use of all the data and to generate improvements in healthcare. When used wisely, AI can bring great accomplishments, but severe restrictions need to be applied in order to make sure that AI will not become a dangerous technology.
AI is a technology involving different matters such as machine learning, deep learning and neural networks. Deep Learning is a technique for implementing machine learning and it uses Artificial Neural Networks that are inspired in our brain and in the biological interconnection between neurons.
How it works? Data Scientists “teaches” machines by inserting endless amounts of data, and the machine will “learn” to recognise specific patterns. It is easy to understand when thinking about imaging recognition. Infants learn what a cat is by showing them pictures of a cat several times. Thus, they can recognise any type of cat as a cat. No matter what colour or shape, a child knows that it is a cat. It works the same for deep learning. After feeding the computer with thousands of cats’ figures, it will be able to recognise any cat, as a cat. Now imagine using this method to recognise a disease in a patient. It can save time and money spent in healthcare.
For example, CDC reports shows that 9.4% of the U.S. population has diabetes. Those patients have high incidence of diabetic retinopathy (DR) and it is a great cause of blindness in the population. Diabetes patients must have a retinal screening periodically. It costs a lot of money and time from the specialists. If we would implement AI in this system, it could save vast amounts of time and money, because only those patients who tested positive for the disease will be sent to an eye specialist for further treatment.
In January 2018, EYE published a study about a new technology that was developed using AI (1). Patients can take a picture of their eyes with the assistance of a smartphone, and the picture will be analysed by an automated AI DR screening. After this, a result, with a high accuracy, will be given on whether you have a disease or not. This technology can especially be helpful in a less developed country with only a small number of specialists available. However, it is important to realise that AI is a tool to help doctors in their work, but doctors are and should be the most important piece in this scene.
AI has a lot of potential to be helpful in healthcare, but the access of data that makes the use of AI possible also brings risks. It is important to have severe rules and regulations for the use of data in order to protect the privacy of the patients.
Imaging recognition is just a small example of what AI can do for us. For sure it will be more and more part of our lives and it is good to understand what it means for our society in respect of healthcare improvement.
1 Ramachandran Rajalakshmi, Radhakrishnan Subashini, Ranjit Mohan Anjana, Viswanathan Mohan. “Automated diabetic retinopathy detection in smartphone-based fundus photography using artificial intelligence. EYE, 32-1138-1144.
Read the Elaine’s first blog here.
We want to thank Elaine, for writing another interesting blog! Do you want to share your story with us, or if you want to participate to the community in any other way, please contact us via info@biotechnologycommunity.com or +31237548660.