Int J Comput Assist Radiol Surg
January 2025
Purpose: This study aims to address the challenging estimation of trajectories from freehand ultrasound examinations by means of registration of automatically generated surface points. Current approaches to inter-sweep point cloud registration can be improved by incorporating heatmap predictions, but practical challenges such as label-sparsity or only partially overlapping coverage of target structures arise when applying realistic examination conditions.
Methods: We propose a pipeline comprising three stages: (1) Utilizing a Free Point Transformer for coarse pre-registration, (2) Introducing HeatReg for further refinement using support point clouds, and (3) Employing instance optimization to enhance predicted displacements.
Int J Comput Assist Radiol Surg
January 2025
Purpose: Lung fissure segmentation on CT images often relies on 3D convolutional neural networks (CNNs). However, 3D-CNNs are inefficient for detecting thin structures like the fissures, which make up a tiny fraction of the entire image volume. We propose to make lung fissure segmentation more efficient by using geometric deep learning (GDL) on sparse point clouds.
View Article and Find Full Text PDFKeratoconus is a burden to health systems and patients worldwide. Corneal collagen crosslinking (CXL) treatment has been shown abroad to be cost-effective for treating progressive keratoconus. However, no cost-effectiveness studies have been performed in Brazil.
View Article and Find Full Text PDFProtein-protein interactions involving 14-3-3 proteins regulate various cellular activities in normal and pathological conditions. These interactions have mostly been reported to be phosphorylation-dependent, but the 14-3-3 proteins also interact with unphosphorylated proteins. In this work, we investigated whether phosphorylation is required, or, alternatively, whether negative charges are sufficient for 14-3-3ε binding.
View Article and Find Full Text PDF