Digital applications of health data have had a rocky beginning when it comes to solving the world’s medical problems. There have been missteps in de-identifying patient data and the sharing of health information without informed consent. In recent weeks, there have been claims made that the benefits of AI to healthcare, in particular reference to IBM’s Watson Health program, have not resulted in enough progress.
Despite the setbacks, there is evidence that the data being harvested, and indeed being given freely by citizens, can make a real difference. With the use of Watson, for example, physicians and hospitals are reporting reduction times with clinical trial matching and changing treatment recommendations for patients.
Alphabet’s artificial intelligence company DeepMind this week reported that using AI to spot signs of eye disease is as effective, and faster, than using experts – resulting in quicker treatment and a better chance of saving a patient’s sight. Similarly, studies have found that machine learning is faster, more accurate and more efficient in analysing echocardiogram heart images than trained experts.
The Internet of Things and big data, as seen in the below infographic, are producing phenomenal amounts of data – up to 2,314 exabytes (1 billion gigabytes) by 2020 – more than two and a half times the capacity the world is expected to have room for by then. It’s hard to imagine that such volumes of information won’t result in valuable insights when examined by machines capable of distilling large amounts of data.
Mistakes will be made along the way. It’s imperative, therefore, that we – tech companies, data companies, healthcare companies, government and citizens – learn from them and make conscious decisions to do better.
That way we’ll hopefully feel better in the future.