Big Data and Industrial Research in Healthcare

Medicine has always been a clinical science supported by observations and data, but for certain aspects medicine is becoming a data science supported by clinicians. Clinicians should be prepared to handle data collected either horizontally from a large number of individuals, or vertically from granular, high resolution, multi-parameters analyses of a single individual (or few). In both cases all the caveats and challenges of big data hold and are similar to those encountered in the Internet of Things (IoT) domain. IoT is described as a network of electronic devices equipped with software, sensors, and connectivity used to collect data for many purposes. By 2020 40% of IoT devices will be related to medicine and health and a huge number of heterogeneous health data will be available beside the already complex datasets produced by omics technologies such as next generation sequencing. This scenario is typical of life science, where high data heterogeneity is both a challenge and an opportunity for data integration aimed at obtaining actionable results exploiting their medical, clinical, predictive values. The long term aim is to be able to measure everything, inside the human body and outside it: a recent example is the study carried out by Mike Snyder's team in Stanford, in which 250,000 measurements were taken daily from 43 individuals for a total of almost 2 billion health data points. Monitoring, elaboration and integration of those data were effectively picking up infections before they actually happened and helped distinguish participants with insulin resistance, a precursor for Type 2 diabetes. Other projects by both academia and the private sector are designed based on the same approach but on a much larger scale, from thousands to millions of monitored volunteers.

The passage to the new century was marked by the delivery of the Human Genome: since then data took the lead in biology and medicine. Medicine gradually acquired new adjectives and characteristics, such as personalized medicine, precision medicine and now high definition medicine. The latter is the ability to assess human health in hi-definition. Since this high granularity is enabled by many new and diverse technologies (NGS applications, sensors monitoring personal physiology and parameters, quantified behavior and lifestyle, advanced imaging) is common to face today a highly heterogeneous flow of data requiring big data capabilities to be integrated.

Multiple and precise measures over time in healthy or diseased individual fits into the "quantified self" paradigm: the habit of self-monitoring during normal activities, trainings, or in the progress of a disease or therapy. It became more and more common to compare the self vs. the self (comparing "you to you") in a practice of quantified self-knowledge that can transform disease prevention putting the patient at the center of the action. With a similar transforming empowerment, a Harvard-MIT startup, NextGen-Jane, aims at a more efficient prevention of the vast number of women's health issues that go undetected. NextGen-Jane analyzes and digitizes menstrual blood as a rich biological matrix to draw a large number of informative data from, tackling hormone levels, fibroids conditions, the vaginal microbiome, fungi or bacterial infections potentially causing cancer.