Neural implicit surface reconstruction of freehand 3D ultrasound volume with geometric constraints

• Published in: Medical image analysis Vol. 98; p. 103305
• Main Authors: Chen, Hongbo, Kumaralingam, Logiraj, Zhang, Shuhang, Song, Sheng, Zhang, Fayi, Zhang, Haibin, Pham, Thanh-Tu, Punithakumar, Kumaradevan, Lou, Edmond H.M., Zhang, Yuyao, Le, Lawrence H., Zheng, Rui
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 01.12.2024
• Subjects: Algorithms; Deep Learning; Freehand 3D ultrasound; Humans; Imaging, Three-Dimensional - methods; Implicit neural representation; Male; Neural Networks, Computer; Phantoms, Imaging; Prostate - diagnostic imaging; Self-supervised surface reconstruction; Signed distance function; Ultrasonography - methods; Self-supervised surface reconstruction; Signed distance function; Implicit neural representation; Freehand 3D ultrasound
• ISSN: 1361-8415; 1361-8423; 1361-8423
• DOI: 10.1016/j.media.2024.103305

Three-dimensional (3D) freehand ultrasound (US) is a widely used imaging modality that allows non-invasive imaging of medical anatomy without radiation exposure. Surface reconstruction of US volume is vital to acquire the accurate anatomical structures needed for modeling, registration, and visualization. However, traditional methods cannot produce a high-quality surface due to image noise. Despite improvements in smoothness, continuity, and resolution from deep learning approaches, research on surface reconstruction in freehand 3D US is still limited. This study introduces FUNSR, a self-supervised neural implicit surface reconstruction method to learn signed distance functions (SDFs) from US volumes. In particular, FUNSR iteratively learns the SDFs by moving the 3D queries sampled around volumetric point clouds to approximate the surface, guided by two novel geometric constraints: sign consistency constraint and on-surface constraint with adversarial learning. Our approach has been thoroughly evaluated across four datasets to demonstrate its adaptability to various anatomical structures, including a hip phantom dataset, two vascular datasets and one publicly available prostate dataset. We also show that smooth and continuous representations greatly enhance the visual appearance of US data. Furthermore, we highlight the potential of our method to improve segmentation performance, and its robustness to noise distribution and motion perturbation. •A self-supervised surface reconstruction method for freehand 3D ultrasound imaging.•Dual novel geometric constraints optimize the signed distance functions end-to-end.•Adaptable to various anatomical structures and robust against noisy inputs.•Enhanced visual quality with smooth, continuous shape representations.