Computerised axial tomography (CT) is an imaging modality currently in widespread use to accurately describe the precise anatomical relationships of soft internal tissues to skeletal structure. A set of CT slices (~ 50) are usually acquired for each patient’s torso. This set of CT images can then be used to construct the patient’s three-dimensional (3D) torso mesh by segmentation. In electrocardiology the diagnosis of myocardial ischaemia and infarction is dependent upon the interpretation of electrical signals measured on the body surface at the time of presentation to hospital. As the electrical current moves from the cardiac surface to the body surface, it is smoothed by the contents of the thoracic cavity. Current algorithms to interpret this surface signal incorporate the dimensions of a standard torso size and shape for an average adult male1 (model torso) which obviously has its limitations. Thus, it is required to develop an scaling algorithm2 which can account for variations in the torso size and shape of each patient, as an approach to further improve the sensitivity and specificity of the algorithm for the diagnosis of an acute coronary syndrome using calculated epicardial potentials.3,4
To develop and evaluate a scaling algorithm operating on a standard torso model (male and female) to generate the tailored human torso mesh based upon few body metrics easily taken from a particular subject under study. The scaled torso mesh will be suitable to derive tailored transfer matrices using boundary element method (BEM) or finite element method (FEM) in the inverse problem to calculate epicardial potentials from body surface potentials.
Using a standard human torso model as the seed torso, which is already available from CT images (DICOM files) construction, a scaling algorithm is required to geometrically reshape the seed torso into a tailored torso for a particular subject (male and female) based upon five specific body linear measurements. To assess the performance of the scaling algorithm, a total of 80 patients torso meshes (40 female and 40 male, with their respective 5 body measurements data) constructed from CT images will be compared with the torso generated by the scaling algorithm using the corresponding body measurements data in each patient. A torso error tolerance evaluation method will be developed and evaluated. The transfer matrix for each scaled torsos mesh (N= 80) will be computed using either BEM or FEM. This will enable the calculation of epicardial potentials from body surface electrocardiographic potentials more accordingly to each patient's body shape and size, hence the possible further improvements on the sensitivity for detecting acute myocardial infarction.4
 “Inverse Epicardial ECG Mapping”, Cesar O. Navarro, PhD Thesis, University of Ulster, 2003.
 “Advanced Methodologies for Diagnosis of Cardiac Conditions”, Colin H. Jamison, PhD Thesis, University of Ulster, 2007.
 “Use of calculated epicardial potentials improves significantly the sensitivity of a diagnostic algorithm in detection of acute myocardial infarction”. C.O. Navarro, C. Owens, J. Riddell, A. McClelland, J. McC. Anderson, O. Escalona, C. Turner and A.A.J. Adgey. Journal of Electrocardiology, Vol. 36, Supp. 1, pp. 127-132, 2003.
 “Improved Detection of Acute Myocardial Infarction Using a Diagnostic Algorithm Based on Calculated Epicardial Potentials”. Colum Owens, Cesar Navarro, Anthony McClelland, John Riddell, Omar Escalona, John McC Anderson, Jennifer Adgey. International Journal of Cardiology, Volume 111, Issue 2, pp. 292-301, 2006.
First Supervisor: Escalona, OJ Prof
Second Supervisor: Finlay, DD Dr
Collaborator: Prof Jennifer Adgey (BHSCT)
Collaboration: In collaboration with Belfast Health and Social Care Trust
The project will develop and evaluate a scaling algorithm operating on a standard torso model (male and female) to generate the tailored human torso mesh based upon few body metrics easily taken from a particular subject under study. The scaled torso mesh will be suitable to derive tailored transfer matrices using boundary element method (BEM) or finite element method (FEM) in the inverse problem to calculate epicardial potentials from body surface potentials.