简介: |
Abstract: Healthcare systems in the developed and developing world are changing: ever-increasing quantities of complex, massively multivariate data concerning all aspects of patient care are routinely being acquired and stored, throughout the life of a patient, both in hospitals (via electronic patient records, or e-health) and in the home (via mobile systems, or m-health). This exponential growth in data quantities far outpaces the capability of clinical experts to cope, resulting in a so-called “data deluge” in which the data are largely unexploited.
This lecture describes “big data” machine learning methods that we have developed to exploit the contents of these complex datasets by performing robust, scalable, automated probabilistic inference. Our goal is to improve healthcare outcomes significantly by developing patient-specific models that help clinicians accurately track the condition of patients in real-time, using the large quantities of data available from patient-worn sensors and other healthcare data. The techniques that we have developed are equally useful in modelling time-series data acquired from biomedical images.
Our research is funded by the UK Government and the Wellcome Trust as a priority centre in personalised medicine and patient safety, and the results of our research are used routinely in the care of 10,000s of patients each month within the UK National Health Service.
|
