Big Data and Machine Learning: Shaping the Future of Healthcare

Co-Chairs: Miguel Ángel Armengol de la Hoz (RCC Fellow, Research Associate at Harvard-MIT Division of Health Sciences and Technology) and Rodrigo Octavio Deliberato (Postdoctoral Fellow at Harvard - MIT, Division Health Sciences & Technology), 

As the nature of the medical profession continues to evolve with advancements in technology, it is becoming increasingly evident that medical knowledge alone does not always provide adequate guidance to make the majority of clinical decisions. According to the 2012 Institute of Medicine Committee Report, only 10 - 20 % of clinical decisions are evidence based.

In a fast-evolving world where computer prices have dropped and processing and storage capacities have increased exponentially, the potential of healthcare data science has become infinite. Technology advancement has led to a massive shift toward the digitalization of patient records, and as a result the healthcare field has created an unprecedented collection of health-related data during the process of care over years.

The immense amount of data our hyper-connected world continues to generate each second has found a perfect ally with the emergence of Big-Data. Utilizing new data-science techniques, healthcare research is no longer limited solely to statistical analysis (testing hypotheses with the support of mathematical methods), rather, it can take new form in the thrilling task of generating and storing knowledge within the memory of computers. This process has become known as ‘machine learning’, a specific type of Artificial Intelligence.

This study group focuses on exploiting these cutting-edge techniques with the goal of bridging the gap between research and healthcare to solve existing clinical problems.

During the sessions carried out, it is expected that the multidisciplinary teams comprised of clinicians, researchers, statisticians, engineers and data-scientists are assembled with the purpose of learning and improving the workflow of the following processes:

  • Integrating, normalizing and structuring an open and de-identified high-resolution clinical data repository.
  • Using innovative automated techniques to mine data in order to answer clinical questions.
  • Promoting and boosting the use and integration of open clinical data repositories through datathons and scientific publications.
  • Employing the state-of-the-art knowledge generated to solve unmet needs at the bedside developing novel clinical decision support systems.

This study group strongly believes that only an international and multidisciplinary team, working collectively to extract trends from big data sets can bring the revolution that modern healthcare needs most. Because ultimately, integrating healthcare innovation (whether technological or not) in any hospital department must be centered on bringing value to the patient.