Technical
Explorations & R&D.

An exploration into X-ray defect classification using autoencoders and Grad-CAM to help make AI decisions more transparent.

A look at single-stage, anchor-free pose estimation for understanding the spatial orientation of aneurysms in 3D space.

An automated approach for detecting candidate aneurysms by utilizing spherical representations and anchor-free learning.

Working with 3D U-Net architectures for aneurysm segmentation, focusing on strategies to manage class imbalance in medical datasets.

Slicer Module for 3D Visualization of the Optimal Cutting Plane Aligned with the Orientation of a Biomedical Object.

A 3D Slicer plugin built to help radiologists verify and classify AI-detected findings in a more intuitive way.