Research using data mining

Introduction

‘Data mining’ projects aim to learn from every patient treated at The Christie with the ultimate aim of being able to better target (or personalize) treatments to the individual disease and characteristics of newly presenting patients. The first steps towards this goal are to begin analyzing the large amounts of data about diagnosis, treatment, radiotherapy planning and outcomes which are already available at The Christie. Using mathematical models and new developments in computational technologies we can extract ANONYMOUS information from existing Christie databases and clinical systems and use this to understand where patient’s treatments have been successful and where there have been unwanted side effects.

Team Objectives

We aim to use the power of ‘big data’ in the future to:

  • Develop systems to support decision making in clinical practice via the Computer Aided Theragnostics project (ukCAT).
  • Link outcomes (survival benefit and treatment toxicity) to individual patient characteristics, imaging results, blood & tissue biomarkers and treatment.
  • Learn from previously treated patients and provide unbiased, individual predictions of potential help and harm of a particular therapy for each newly presenting patient.
  • Enable this information, in the future, to be used together by patients and their doctors to reach shared, informed decisions about the most appropriate treatment.
  • Extend ‘big data’ analytics to the complex, multi-dimensional clinical and imaging data collected every day during routine cancer treatment in the NHS.

 

;