Show simple item record

dc.contributor.authorZhou, Shijie
dc.date.accessioned2018-01-15T15:24:23Z
dc.date.available2018-01-15T15:24:23Z
dc.date.issued2018-01-15T15:24:23Z
dc.identifier.urihttp://hdl.handle.net/10222/73567
dc.description.abstractThe clinical electrophysiological (EP) techniques of intracardiac recording and stimulation have emerged as invaluable tools for investigating and treating life-threatening cardiac arrhythmias. Computing technology plays a crucial role in making EP techniques possible. Two computational approaches designed to facilitate EP procedures are the subject of this dissertation: the first one is referred to as electrocardiographic imaging (ECGI); the second one is a novel statistical approach that enables a real-time guidance of the EP procedure based on the standard 12-lead ECG and a pace-mapping. Pre-ablation planning by means of ECGI can help to localize the origin of ventricular tachycardia (VT) and thus contribute to improving ablation-procedure outcome. The classical ECGI solves the inverse problem of electrocardiography by reconstructing epicardial potentials from multiple body-surface ECGs and patient-specific geometry of the heart and torso acquired by computed tomography. To overcome the inherent instability of the inverse problem, regularization methods must be used to constrain the solution. The present study assessed a data-driven Bayesian approach to the inverse solution that uses a novel algorithm for deriving dynamic spatio-temporal constrains for the solution. The encouraging results of validation experiments provide a strong incentive for pursuing the Bayesian method further. Next, a new statistical technique for real-time guidance of EP procedure was investigated. This technique supplements electroanatomic mapping---which provides patient's heart geometry---and it requires only the 12-lead ECG for a sufficient number of pacing sites with known coordinates to develop multiple linear regressions for predicting the origin of unknown activation sequence. The localization accuracy of the latter statistical method was superior to that achieved by the inverse solution and thus this approach to localizing the origin of ventricular activation offers an alternative to the pre-procedure inverse solution and its implicity makes it practical for real-time applications.en_US
dc.language.isoen_USen_US
dc.subjectBody-surface potential mappingen_US
dc.subjectAblation-Catheteren_US
dc.subject12-lead ECGen_US
dc.titleLOCALIZATION OF VENTRICULAR ACTIVATION ORIGIN USING PATIENT-SPECIFIC GEOMETRYen_US
dc.date.defence2017-12-11
dc.contributor.departmentDepartment of Biomedical Engineeringen_US
dc.contributor.degreeDoctor of Philosophyen_US
dc.contributor.external-examinerDr. Dana Brooksen_US
dc.contributor.graduate-coordinatorDr. Robert Adamsonen_US
dc.contributor.thesis-readerDr. John L. Sappen_US
dc.contributor.thesis-readerDr. Alexander Quinnen_US
dc.contributor.thesis-supervisorDr. B. Milan Horaceken_US
dc.contributor.thesis-supervisorDr. Joshua L. Leonen_US
dc.contributor.ethics-approvalReceiveden_US
dc.contributor.manuscriptsNot Applicableen_US
dc.contributor.copyright-releaseNot Applicableen_US
 Find Full text

Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record