Skip to Main content Skip to Navigation

Aide à la décision pour le remplacement valvulaire aortique percutané

Abstract : Aortic stenosis represents the most frequent acquired valvular heart disease, affecting up to 10% of octogenarians. Transcatheter aortic valve implantation (TAVI) is booming and confronts clinicians with new issues that constitute a major field of research. Our work falls within the framework of computer-assisted medico-surgical interventions, and aims at proposing computer-assisted decision support systems. The present Thesis is composed of four parts. The first part focuses on the medical problematic surrounding TAVI, as well as the current French TAVI field on the basis of an article describing temporal trends in patients’ and procedural’s characteristics from 2010 to 2015 in the FRANCE 2 and FRANCE TAVI nationwide registries. This first part identifies medical issues that operators currently face, especially the optimal selection of TAVI candidates, and the reduction of procedural complications within the current trends towards treatment of patients with lower baseline surgical-risk profile. The second part deal with population-based studies, through standard statistical methods, to identify predictors of TAVI outcomes or selected procedural complications in order to facilitate procedural planning. Three articles compose this part. The first focuses on predictors of short-term cerebrovascular events post-TAVI, the second deals with conduction disturbances post-TAVI while the third aims at identifying predictors of global poor outcomes. We demonstrate the benefits of these analyses, which will remain necessary in the future, but also address their limitations, which support the use of new methods to store, sort, retrieve, and even augment relevant information to facilitate operators’ decision, especially at the pre-procedural step.The purpose of Part 3 is to address a case-based reasoning (CBR) decision-support system that could benefit from the identification of these prognostic factors and ultimately integrate them into a global and ergonomic interface for decision support. We have worked in the framework of the European project H2020 EurValve on the development of a CBR whose problematic is,for the time being, limited to the optimal choice of the approach, type and size of prosthesis. Our work focused on an analytical step in the design of this type of system dealing with the study and improvement of the similarity measure used to identify nearest neighbours (previously treated cases and their therapeutic "solution") of the current problem (case which clinicians are planning to treat). Finally, the last part focuses on increasing the information available for preoperative decision support through patient-specific numerical simulation. After a state of the art of the methods used in the field of TAVI, we worked on the elaboration and parameterization of a simulation model of the insertion of the stiff guidewire in the left ventricle (one of the first steps of the procedure that can condition the positioning of the prosthesis and thus the final result). In order to perform a first validation of this patient-specific simulation using preoperative 3D CT imaging, the proposed approach is based on the extraction of the region of interest in the 3D volume (segmentation) and its mapping to intraoperative 2D fluoroscopy through 3D / 2D registration. Our work on these image processing methods needed to implement and validate our simulation strategy is also discussed in this section. Finally, we present a potential clinical application of the simulation model regarding the influence of the shape of the guide and its insertion conditions on its stability and the pressure forces exerted on the left ventricle.
Document type :
Complete list of metadata

Cited literature [176 references]  Display  Hide  Download
Contributor : ABES STAR :  Contact
Submitted on : Thursday, October 8, 2020 - 1:02:43 AM
Last modification on : Wednesday, September 14, 2022 - 10:20:04 AM
Long-term archiving on: : Saturday, January 9, 2021 - 6:05:55 PM


Version validated by the jury (STAR)


  • HAL Id : tel-02960817, version 1


Vincent Auffret. Aide à la décision pour le remplacement valvulaire aortique percutané. Médecine humaine et pathologie. Université Rennes 1, 2019. Français. ⟨NNT : 2019REN1B035⟩. ⟨tel-02960817⟩



Record views


Files downloads