Publications by authors named "A Kaladji"
Article Synopsis
- The study presents a new weakly-supervised deep learning framework for automatically segmenting vascular structures in non-contrast CT angiography images, addressing limitations of traditional strong-label methods.
- It introduces an efficient annotation process that uses pseudo labels generated through 2D Gaussian heatmaps, significantly reducing the time needed for labeling by about 82%.
- Results from various datasets reveal that this new method outperforms fully-supervised learning in segmentation accuracy (up to a 2.74% increase in Dice score) while allowing the incorporation of additional unlabeled data.
Background: The aim of this study was to evaluate a new measurement tool for the preoperative computed tomography (CT) scan to estimate the endograft apposition surface on the proximal aortic neck of the aneurysm that would predict the risk of late type IA endoleak (EL1A).
Methods: From databases of 4 high-volume centers between 2004 and 2020, all patients who underwent Endovascular Aneuruysm Repair (EVAR) with the possibility of analyzing pre and post-EVAR scans and who presented late (>2 years) EL1A in follow-up were included retrospectively (EL1A group). A control group of randomly selected EL1A-free patients were included (control group) and followed-up beyond 4 years.
Article Synopsis
- The study compares the hemodynamic effects of four-vessel fenestrated endovascular aneurysm repair (fEVAR) and branched endovascular aneurysm repair (bEVAR) in patients with complex aortic aneurysms using patient-specific computational flow simulation (CFS).
- Patients who underwent either fEVAR or bEVAR were analyzed, revealing significant differences in renal artery flow rates and pressure, with fEVAR providing better renal perfusion outcomes.
- The analysis found that while bEVAR negatively impacted renal artery function, there were no substantial differences in performance for celiac or superior mesenteric arteries between the two techniques.
A parametric study assessing Implicit Solver limits for a generic FEM Simulation of PTA without stent deployment.
Annu Int Conf IEEE Eng Med Biol Soc
July 2023
This study focuses on the robustness of a generic Finite Element Model (FEM) of Percutaneous Transluminal Angioplasty (PTA) procedure with permanent set. The influence of three different parameters on simulation robustness were investigated: the stenosis percent, the stenosis offset and the arterial caliber. Five arterial calibers are modeled by adapting the ratio between the inner diameter and the wall thickness.
View Article and Find Full Text PDF