Bosi, GM;
(2016)
Patient-specific computational modelling for transcatheter valve implantation: Towards translation into clinical practice.
Doctoral thesis , UCL (University College London).
Abstract
Non-surgical procedures for heart valve disease are now a reality with transcatheter valve implantation used to treat dysfunction of both the aortic and pulmonary valve. Patient selection, crucial to guarantee successful procedural and long-term outcomes, relies on clinical image investigations. These can be combined with advanced patient-specific computational modelling to obtain additional, predictive information about the response of individual patients to device implantation. However, to be clinically meaningful, computational models have to be fully validated and, in this context, consider not only the morphology, but also the mechanical properties of the individual implantation site. In this thesis, I have analysed cases of both aortic and pulmonary transcatheter valve implantation in order to identify and address current pitfalls of the computational methodology. I have developed a patient-specific finite element (FE) model of transcatheter aortic valve implantation (TAVI) that realistically mimics different devices on the market. I have tested it in a large cohort of patients retrospectively selected and validated it against clinical data. TAVI patients present a rigid calcified implantation site; therefore, the mechanical properties of the artery have less influence on the results. On the contrary, in percutaneous pulmonary valve implantation (PPVI), the implantation site is highly dynamic. Thus, I have applied the methodology developed for the FE model of TAVI to a case of PPVI where the patient presented borderline criteria for a conventional approach, in order to help select the optimal treatment strategy. I have focused on the implantation site mechanical response, which has driven a set of experimental studies for ballooning procedures. Reverse engineering methodology can be used to derive the implantation site overload response (i.e. during percutaneous procedure) during ballooning for more accurate computational modelling. Increasingly refined patient-specific computational models could be used to help procedural planning and predict post-operative outcomes, thus enhancing patient selection.
Archive Staff Only
 |
View Item |
