shuai zhang
@ucl.ac.uk
Department of Computer Science
university college london
RESEARCH, TEACHING, or OTHER INTERESTS
Engineering, Artificial Intelligence
Scopus Publications
Scholar Citations
Scholar h-index
Scholar i10-index
Scopus Publications
Kai Pan, Shuai Zhang, Liang Zhao, Shoudong Huang, Yanhao Zhang, Hua Wang, and Qi Luo
IEEE
The 3D reconstruction of patient specific bone models plays a crucial role in orthopaedic surgery for clinical evaluation, surgical planning and precise implant design or selection. This paper considers the problem of reconstructing a patient-specific 3D tibia and fibula model from only two 2D X-ray images and one 3D general model segmented from the lower leg CT scans of one randomly selected patient. Currently, the bone 3D reconstruction mainly relies on computed tomography (CT) and magnetic resonance imaging (MRI) scanning-based mode segmentation which result in high radiation exposure or expensive costs. While, the proposed algorithm can accurately and efficiently deform a 3D general model to achieve a patient-specific 3D model that matches the patient's tibia and fibula projections in two 2D X-rays. The algorithm undergoes a preliminary deformation, 2D contour registration, and opti-misation based on the deformation graph that represents the shape deformation of models. Evaluations using simulations, cadaver and in-vivo experiments demonstrate that the proposed algorithm can effectively reconstruct the patient's 3D tibia and fibula surface model with high accuracy.
Shuai Zhang, Liang Zhao, Shoudong Huang, Hua Wang, Qi Luo, and Qi Hao
Springer Nature Switzerland
Shuai Zhang, Liang Zhao, Shoudong Huang, Menglong Ye, and Qi Hao
Institute of Electrical and Electronics Engineers (IEEE)
This article presents a framework for 3D reconstruction of colonic surface using stereo colonoscopic images. Due to the limited overlaps between consecutive frames and the nonexistence of large loop closures during a normal screening colonoscopy, the state-of-art simultaneous localization and mapping (SLAM) algorithms cannot be directly applied to this scenario, thus a colon model segmented from CT scans is used together with the colonosocopic images to achieve the colon 3D reconstruction with high accuracy. The proposed framework includes 3D scan (point cloud with RGB information) reconstruction from stereo images, a visual odometry (VO) based camera pose initialization module, a 3D registration scheme for matching texture scans to the segmented colon model, and a barycentric-based texture rendering module for mapping textures from colonoscopic images to the reconstructed colonic surface. A realistic simulator is developed using Unity to simulate the procedures of colonoscopy and used to provide experimental datasets in different scenarios. Experimental results demonstrate the good performance of the proposed 3D colonic surface reconstruction method in terms of accuracy and robustness. Currently, the framework requires a pre-operative colon model as the template for colon reconstruction and can reconstruct 3D colon maps when the colon has no large deformation and the colon structure is clearly visible. The datasets used in this article and the developed simulator are made publicly available for other researchers to use (https://github.com/zsustc/colon_reconstruction_dataset).
Shuai Zhang, Liang Zhao, Shoudong Huang, Ruibin Ma, Boni Hu, and Qi Hao
IEEE
In colonoscopy procedures, it is important to rebuild and visualize the colonic surface to minimize the missing regions and reinspect for abnormalities. Due to the fast camera motion and deformation of the colon in standard forward-viewing colonoscopies, traditional simultaneous localization and mapping (SLAM) systems work poorly for 3D reconstruction of colon surfaces and are prone to severe drift. Thus in this paper, a preoperative colon model segmented from CT scans is used together with the colonoscopic images to achieve the 3D colon reconstruction. The proposed framework includes dense depth estimation from monocular colonoscopic images using a deep neural network (DNN), visual odometry (VO) based camera motion estimation and an embedded deformation (ED) graph based non-rigid registration algorithm for deforming 3D scans to the segmented colon model. A realistic simulator is used to generate different simulation datasets with ground truth. Simulation results demonstrate the good performance of the proposed 3D colonic surface reconstruction method in terms of accuracy and robustness. In-vivo experiments are also conducted and the results show the practicality of the proposed framework for providing useful shape and texture information in colonoscopy applications.
Shuai Zhang, Ruihua Han, Wankuan Huang, Shuaijun Wang, and Qi Hao
IEEE
In this paper, we propose an improved UWB based indoor localization system using Bayesian filtering techniques. The system contains two key components: (1) miniaturized, high updating rate and highly reconfigurable UWB sensors with a linear regression model to calibrate range measurement errors; (2) a set of Bayesian filters which can improve the localization precision by utilizing the spatial correlation between the stationary UWB base stations and the mobile UWB station. Furthermore, a novel measurement transform is proposed to reduce the computational complexity. Experiments are performed in an indoor environment with the ground truth obtained by the motion capture system to validate and evaluate the proposed indoor localization system.
Shuai Zhang, Shuaijun Wang, Chengyang Li, Guocheng Liu, and Qi Hao
IEEE
The autonomous navigation of unmanned aerial vehicles (UAVs) require a lot of sensing modalities to improve their cruise efficiency. This paper presents a system for autonomous navigation and path planning of UAVs in GPS-denied environment based on the fusion of geo-registered 3D point clouds with proprioceptive sensors (IMU, odometry and barometer) and the 2D Google maps. The contributions of this paper are illustrated as follows: 1) combination of 2D map and geo-registered 3D point clouds; 2) registration of local point cloud to global geo-registered 3D point clouds; 3) integration of visual odometry, IMU, GPS and barometer. Experiment and simulation results demonstrate the efficacy and robustness of the proposed system.