Integrated Modelling of Functional Imaging and Gene Expression Data for Personalised Cancer Therapy

Cancer is one of the primary causes of death in the world. In the United Kingdom, one in four people will die of cancer, whilst one in three will be diagnosed with cancer at some point in their life. The effective treatment of cancer is therefore a topic of considerable importance to both medical practitioners and the general public. The proposed research aims to improve the treatment of cancer by combining advanced methods from information engineering and computer science with modern imaging and genomic technologies, in order to enable personalised treatment options for individual cancer patients.At the present time, most cancer patients are treated according to the general type of cancer from which they suffer (e.g. breast cancer, colorectal cancer etc.). A significant problem with this approach however is that not every patient responds equally well to the same course of therapy. In the case of colorectal cancer for example, approximately 20% of patients do not respond successfully to chemoradiotherapy (combined chemotherapy and radiation therapy). As a result, many colorectal cancer sufferers undergo an unnecessary course of this treatment, experiencing the significant side-effects of chemotherapy in the process without ultimately gaining any net benefit from the treatment itself.Personalised cancer therapy is an important and emerging area of research which offers the potential for a dramatic improvement in the treatment of cancer. The aim of this field is to develop techniques which enable specific, individualised treatment options for cancer patients. Thus, for a particular patient with a given tumour, personalised therapy involves determining the specific course of treatment (from the wide range of options available) which will be most beneficial for the individual patient.Although the field of personalised cancer therapy offers much promise for the future treatment of cancer, there are a number of key scientific challenges to be addressed before this promise can be realised. Most importantly, it is necessary to develop robust methods for predicting treatment outcome based on an analysis of the biological characteristics of the specific tumour under consideration. In order to address this problem, the proposed research will make use of two state-of-the-art techniques for assessing the intrinsic biological properties of a tumour, namely functional imaging and gene expression profiling .Functional imaging techniques such as functional computed tomography (functional CT) and positron emission tomography (PET) can be used to image the functional or physiological characteristics of a tumour (such as the degree of blood flow to the tumour, or the level of metabolism within the tumour). Conversely, gene expression profiling of a given tumour involves the accurate measurement of the expression levels of the different genes within a particular group of tumour cells. Given the complementary nature of functional imaging and gene expression profiling, the combination of these two techniques offers the potential to provide a more effective profiling of cancer, compared with that available from either technique alone.The aim of the proposed research is to develop new algorithms for integrating functional imaging and gene expression data in order to provide an improved prediction of the likely response to therapy. To achieve this aim, the research will make use of modern statistical methods from the fields of information engineering and computer science. Such methods provide an effective means of combining different types of information, and will enable the development of powerful new algorithms for the intelligent analysis of cancer imaging and genomic data. These algorithms will provide both clinicians and cancer researchers with a new tool to predict the effectiveness of different treatment options for individual cancer patients.