A sedentary lifestyle, poor diet, smoking, and genetic and other health factors are major contributors to coronary heart disease (CHD). Despite recent medical advances that have lowered the number of deaths compared to the past decades, CHD still remains the number 1 disease in mortality in the UK (73,000 deaths per year) with a tremendous economic burden: estimates put the cost to UK's economy at £6.7 billion per year. The overriding goal of this project is to take advantage of multimodal information within cardiac magnetic resonance images to improve their analysis and facilitate the diagnosis and improve treatment of CHD. Magnetic Resonance Imaging (MRI) as an imaging diagnostic tool is uniquely positioned to help as it is non-invasive and does not use radiation. A typical cardiac protocol relies on several MR imaging sequences to provide images of different contrast, termed as modalities hereafter, to assess disease progression and status. As a result of this range of acquisitions, hundreds of multidimensional multimodal images are generated in a single patient exam leading to severe data overload. Therefore, robust and automated analyses algorithms would help alleviate the clinical reading burden. Several algorithms have been proposed to segment and register the myocardium in the most commonly used modalities by considering them independently. However, the problem remains difficult and performance is not yet adequate. Currently, the analysis of cardiac imaging data still remains a manual, time consuming, and expensive process typically performed by clinical experts. As a result, despite the huge amount of data generated, not only in a clinical but also in a research setting, only a fraction is being analysed robustly, due to the vast amount of time required for the analysis of this data. This proposal aims to address the above shortcomings by proposing mechanisms that take advantage of the shared information that exists across modalities to enable the joint analysis of cardiac imaging data and thus make a significant leap in how we approach their analysis. We propose new multimodal machine learning driven mechanisms to learn image features (i.e. how local image information is represented for an algorithm to use) that do not change between imaging modalities whilst preserving shared anatomical information. We will then use the learned features in multimodal patch-based myocardial segmentation and inter-modality non-linear registration (i.e. the non-linear registration between two images coming from different cardiac MR sequences) thus enabling us to relate images of the same patient across different modalities. To maximise impact, we will develop an inter-modality cardiac registration plugin for a commercial clinical package that is also offered as an open source variant for academic purposes. We expect that when our complete framework is integrated into clinical tools and becomes widely available it can radically change current clinical reading workflow and decision-making. It will permit the propagation of annotations across multimodal images of a patient exam effortlessly and seamlessly, thus significantly reducing reading time and permitting the analysis of cardiac data on a larger scale.

In the diagnosis and treatment of disease, clinicians base their decisions on understanding of the many factors that contribute to medical conditions, together with the particular circumstances of each patient. This is a "modelling" process, in which the patient's data are matched with an existing conceptual framework to guide selection of a treatment strategy based on experience. Now, after a long gestation, the world of in silico medicine is bringing sophisticated mathematics and computer simulation to this fundamental aspect of healthcare, adding to - and perhaps ultimately replacing - less structured approaches to disease representation. The in silico specialisation is now maturing into a separate engineering discipline, and is establishing sophisticated mathematical frameworks, both to describe the structures and interactions of the human body itself, and to solve the complex equations that represent the evolution of any particular biological process. So far the discipline has established excellent applications, but it has been slower to succeed in the more complex area of soft tissue behaviour, particularly across wide ranges of length scales (subcellular to organ). This EPSRC SoftMech initiative proposes to accelerate the development of multiscale soft-tissue modelling by constructing a generic mathematical multiscale framework. This will be a truly innovative step, as it will provide a common language with which all relevant materials, interactions and evolutions can be portrayed, and it will be designed from a standardised viewpoint to integrate with the totality of the work of the in silico community as a whole. In particular, it will integrate with the EPSRC MultiSim multiscale musculoskeletal simulation framework being developed by SoftMech partner Insigneo, and it will be validated in the two highest-mortality clinical areas of cardiac disease and cancer. The mathematics we will develop will have a vocabulary that is both rich and extensible, meaning that we will equip it for the majority of the known representations required but design it with an open architecture allowing others to contribute additional formulations as the need arises. It will already include novel constructions developed during the SoftMech project itself, and we will provide many detailed examples of usage drawn from our twin validation domains. The project will be seriously collaborative as we establish a strong network of interested parties across the UK. The key elements of the planned scientific advances relate to the feedback loop of the structural adaptations that cells make in response to mechanical and chemical stimuli. A major challenge is the current lack of models that operate across multiple length scales, and it is here that we will focus our developmental activities. Over recent years we have developed mathematical descriptions of the relevant mechanical properties of soft tissues (arteries, myocardium, cancer cells), and we have access to new experimental and statistical techniques (such as atomic force microscopy, MRI, DT-MRI and model selection), meaning that the resulting tools will bring much-need facilities and will be applicable across problems, including wound healing and cancer cell proliferation. The many detailed outputs of the work include, most importantly, the new mathematical framework, which will immediately enable all researchers to participate in fresh modelling activities. Beyond this our new methods of representation will simplify and extend the range of targets that can be modelled and, significantly, we will be devoting major effort to developing complex usage examples across cancer and cardiac domains. The tools will be ready for incorporation in commercial products, and our industrial partners plan extensions to their current systems. The practical results of improved modelling will be a better understanding of how our bodies work, leading to new therapies for cancer and cardiac disease.