The combination of big data and artificial intelligence (AI) is a powerful one that is proving to be a game changer for many industries. Until recently, though, it has had relatively little impact on the healthcare sector, where the data concerned is particularly sensitive and there are stringent controls on how it is used.
Assuming that appropriate controls and security measures are in place, there is a huge opportunity here – a recent Fortunearticle estimated that around 30% of the world's data production is health-related. That includes everything from laboratory tests, medical images, genetic profiles, liquid biopsies and electrocardiograms to data from insurance claims, clinical trials, prescriptions and academic research. This data has been collected for years, but it's only now that the technology exists to make good use of it.
This has been noted at the highest levels; for instance, in a recent speech, UK Prime Minister Theresa May urged the National Health Service and technology companies to use AI as a "new weapon" in research, saying that she wants to see computer algorithms sifting through patients' medical records, genetic data and lifestyle habits to spot cancer earlier and enable it to be treated.
That is an ambition for the future, but there are already examples of big data and AI working together to improve efficiency – and, crucially, medical outcomes – in healthcare.
The value of data
One practical use of health data is in streamlining the process of developing new drugs and bringing them to market, and California-based biotechnology company Amgen is making the most of this opportunity.
One of the world's most valuable health databases originated in Iceland, whose geographical isolation makes it a fertile subject for genetic research. In the 1990s, the government oversaw the collection of the genetic sequences of 160,000 Icelandic citizens, along with their medical and genealogical records.
In 2012, Amgen bought the company that was in charge of storing and analyzing that data, and the purchase has transformed its research and development process. Previously, only 15% of Amgen candidate molecules had been validated against specific genetic targets. After the purchase, Amgen began evaluating all of its drug candidates against the Icelandic database and today, three-quarters of its pipeline is based on genetic insights largely gleaned from that database.
On an individual level, smartphones and other connected devices are changing the relationship between patients and their health data, and enabling them to improve their personal health in the process. For example, Apple is working with Stanford University in California on an ongoing study to explore whether wearable devices can detect serious cardiac conditions by identifying irregular heart rhythms.
Similarly, digital diabetes prevention and treatment platforms such as Virta and Omada Health connect users with support communities and health coaches who can remotely monitor factors such as weight, blood sugar, diet and medicine intake.
Another example is the ingestible sensor created by Proteus Digital Health, which helps patients (and, if they wish, their doctors and family members) to keep track of whether or not they're taking their medication. This isn't just a gimmick; encouraging people to take the drugs at the right time not only helps them to recover, but also saves on expensive medication.
Heart disease recovery
On a larger scale, a study published in the Journal of the American Heart Associationby researchers at Yale University reveals how big data can help to predict a patient's chances of surviving a heart attack.
Doctors analyzed the health data of more than 40,000 patients. Using a machine learning statistical technique (a form of AI), they began to predict the condition of patients one year after the heart attack. They also used cluster analysis to sort patients into four distinct categories based on their responses to the most common treatments.
Subsequently, researchers used their findings to develop an online predictive tool thatcan be integrated with electronic health records in healthcare systems. In the long run, their goal is to use analytics to provide personalized care to heart attack victims.
AI technology can also be applied to data at the individual level. For example, in April, the US Food and Drug Administration authorized the marketing of the first medical device to use AI to detect eye disease.
The disease in question is diabetic retinopathy, which occurs when high levels of blood sugar lead to damage in the blood vessels of the retina. It is the leading cause of vision impairment and blindness among adults.
The device, IDx-DR, is a software program that uses an AI algorithm to analyze images of the eye taken with a retinal camera. A doctor uploads the digital images of the patient's retinas to a cloud server, and the software returns a positive or negative result. If it is positive, patients are advised to see an eye care specialist for further evaluation and possible treatment.
IDx-DR is the first device authorized for marketing that provides a screening decision without the need for a doctor to interpret the image or results as well. This means that healthcare providers who may not normally be involved in eye care can use it.
The future of healthcare
It's tempting to conclude that, as data sets grow bigger, computers get faster and algorithms get smarter, AI will eventually remove the need for human doctors entirely. As with most predictions of the technological future, this is unlikely to occur: human judgement and skill will always be at the center of healthcare. However, AI is set to play an increasingly important role in maximizing the efficiency and effectiveness of that care.