Trackerless 3D Point-Of-Care Ultrasound Reconstruction

Reconstructing 2D ultrasound (US) images into 3D volumes offers significant advantages, from quantitative measurements to enhanced 3D visualization and interventional guidance. Trackerless 3D freehand ultrasound reconstruction further enhances diagnostic capabilities by providing 3D imaging without the need for external tracking systems, making it more accessible and easier to use in clinical settings.

Over the past year, we have been working on a novel approach for trackerless 3D ultrasound reconstruction at The Netherlands Cancer Institute, Image-Guided Surgery group. This research utilizes deep learning (CNN-Transformer) combined with IMU data integrated into the point-of-care ultrasound. This work represents a significant step toward making advanced 3D US imaging more accessible and effective in clinical practice, with potential applications in diagnostic imaging, surgical planning, and intraoperative guidance