Artificial vs Biological intelligence in healthcare

Ranging from electronic health records to public health datasets, today, healthcare generates a vast amount of digital information. Such a wealth of data presents an exciting opportunity for integrated machine learning solutions to address problems across multiple facets of healthcare practice and administration. To help in making decisions and to extract useful knowledge we utilise in this thesis predictive analytics, machine learning and statistical modeling. We start by exploring pre-surgery patients profiles, capturing patterns and revealing associations between data as well as statistically significant differences. We then use that findings for building pre-operative predictive models for patients undergoing knee or hip surgery. We focus our study in two main research questions, starting with the prediction of patient clinical milestones, where we are interesting in building regression models that can accurately predict the day that a patient will reach a certain milestone. In the second research question we turn our attention to the Patient Reported Medical Outcomes (PROMs) where a range of classification models have been built, predicting an interval range associated with the medical score of each patient. Both of our research questions contribute to the medicine and personalised treatment, helping the healthcare specialists to have an expectation in advance about the pace of rehabilitation of their patients.