Big data. Quite possibly the buzzwords of this decade. For the first time in history, our increasingly interconnected world has been allowing industries of all kinds to amass incredible amounts of data about us all. From Google searches, to places we visit, to the posts we like on Facebook, and beyond. Much of this data has been leveraged to advertise to us in more and more effective and targeted ways. That’s great for marketers and the companies that hire them, but there are other far greater benefits to humanity just around the corner thanks to big data.
Healthcare gathers huge amounts of data already; digitised patient records are just the tip of the iceberg, with rapid advances in genomics and related disciplines generating mind-boggling amounts of data. With the cost of a full genome sequence recently ducking below the $1,000 mark, it has been predicted that, by 2025, the data generated annually by the field of genomics alone will be equivalent to what is generated by astronomical science, YouTube and Twitter combined!
When you combine these huge datasets with associated patience records and outcomes, as well as the large amounts of research being performed in every college and university across the globe, the challenge is how to make use of all of this. That’s where machine and deep learning come in. Machine learning is one key field of artificial intelligence, and deep learning is a subset of this that seeks to train neural networks to analyse and identify patterns in information
Indeed, there are many players in the field already, with one of the most high-profile ones being IBM’s Dr Watson. Dr Watson is currently providing assistance to oncologists, and has advanced capabilities with regards to analysing clinical notes. It takes this data as well as external research and data in order to provide the clinician with potential treatment plans. Another example is Google’s Deepmind Health project, which is a pilot project that mines data from medical records in order to provide improved health outcomes – in this case specifically with Moorfields Eye Hospital NHS Foundation Trust to improve eye-related treatment.
With regards to genomics and the huge associated data sets, deep learning techniques are being leveraged in order to identify patterns between these and patient medical records en masse, looking for potential mutations that may be associated with disease, future risks, or responsiveness to drugs. This will allow doctors to monitor patients more effectively for diseases they may be at risk of, to more effectively treat patients when they do get ill, and again all of those outcomes will be fed back into patient data allowing for more further refinement of treatment and diagnoses.
It is clear that the future of healthcare has the potential to be very bright indeed, with deep learning and machine learning giving us the ability to make sense of large swathes of data in a way that has never before been possible. It’s no wonder that so many tech companies are betting big on this field, it seems to be a natural evolution for Silicon Valley. Let’s see where it all goes!