Although the large amounts of data being generated in the health sector are still a long way from providing solutions in themselves, they do nevertheless provide insights into a broad range of cases and may thus make it easier to find effective remedies.
A recent L'Atelier article focused on the Human Face of Big Data project, whose goal is to gather both active and passive data to try to “address the pressing issues of our time.” One of these important issues is health. As part of the project, both entrepreneurs and scientists have presented findings highlighting some of the concrete benefits of Big Data for those working in such fields as diabetes, malaria and heart attacks. They have found that the data they collect either in turn helps gather even more information or to come up with a more precise picture of how treatment or prevention is carried out in different areas. To take an example: combating malaria. One Big Data project takes the suggestions made by Twitter users for combating the spread of the disease and uses a social network analysis system based on the frequency of and links between certain key words. The software identifies from which countries and sources the tweets were sent.
“If we’re careful and wise, in the not too distant future this new set of technologies may have an impact on humanity as great as those of language and art,” insists Rick Smolan, the project founder. Sébastien Sasseville is living proof of this. A type-1 diabetes sufferer, this triathlon competitor uses biometric sensors to help him prepare for these extreme physical challenges and to enable science to help him overcome his condition. Also a mountaineer, Sasseville actually reached the summit of Mount Everest. During the climb he used data from a number of biometric sensors to check his state of health and modulate his efforts. He was able to assess his capabilities on a daily basis by studying his body temperature, insulin level, calorie and salt consumption, plus his GPS coordinates, altitude and heart rate. Furthermore, Brendan Moran, data scientist at EMC Greenplum, a division of EMC², which is sponsoring the Human Face of Big Data project, told us: “the UK government and state-run hospitals in the UK are now able to publish their data on diabetes. Aggregating data will help to validate the research and so to achieve progress.”
Strong data flow boosts ‘machine learning’
In a third example, researchers John Guttag, Collin Stultz and Zeeshan Syed have created a computer model to analyse formerly discarded electrocardiographic (ECG) data from heart attack patients. Using machine learning and data mining to sift the huge volumes of available data, they found three ECG abnormalities that correlate with a higher risk of dying from a second heart attack during the same year. They believe their computer model can significantly improve on traditional risk-screening techniques, which currently fail to identify and anticipate as many as 70% of repeat heart attack cases.