This study addresses the challenge of reconstructing 3D anatomical structures in medical imaging, focusing on ultrasound, which typically provides 2D images. Existing 3D ultrasound technologies require specialized transducers and have limited scan ranges, while freehand 3D reconstruction is not yet applied. This research develops a method using AI to estimate the transducer's movement based on the relationships between consecutive images, overcoming the need for sensors that could interfere with the scanning process. Additionally, by integrating photoacoustic imaging with ultrasound, the study demonstrates the 3D reconstruction of microvascular structures. This innovation advances ultrasound and photoacoustic imaging, enabling more accurate anatomical analysis and expanding 3D visualization to complex areas like blood vessels, nerves, organs, and tumors.

- Authors (Pusan National University) 

 · SiYeoul Lee (Department of Information Convergence Engineering)

 · MinWoo Kim (School of Biomedical Convergence Engineering)

- Title of original paper: Enhancing Free-hand 3D Photoacoustic and Ultrasound Reconstruction using Deep Learning

- Journal: IEEE Transactions on Medical Imaging

- Web link: https://ieeexplore.ieee.org/document/11036110 

- Contact e-mail: mkim180@pusan.ac.kr