No abstract

Endovascular stent-graft implantation technique is a newly developed, effective and less invasive method in treating thoracic aortic dissection (TAD). Our study was designed to further verify the feasibility, the efficacy, and safety of this technique. We present a 4-year follow-up report of endovascular stent-graft treatment over 36 cases of acute TAD patients and 40 cases of chronic TAD patients. The mortality and comorbidity rates were evaluated thoroughly. In our study, the deployment of the stent-grafts was successfully performed in 75 cases. The hospital cumulative 30-day mortality rate was 1.3%. The instant endoleak rate was 15.8% (12 patients). All endoleaks were successfully treated with a second stent. All patients in local anesthesia were transported to the general ward after the intervention and were discharged from hospital within 1 week. Our preliminary results showed endovascular stent-graft implantation technique offered good peri-operative morbidity and mortality rates. Stent-graft placement over TAD produced a low incidence of spinal cord ischemia, cardiac and pulmonary complications, less hospital stay, less blood transfusion and became the first choice of TAD patients in our department.

Input Results Side ViewBird's-eye ViewFigure 1. Given a single large-scene image with hundreds of people, our method can reconstruct 3D poses, shapes and locations of these people in a global camera space with coherency with the scene. Please zoom in for more details.

3D face reconstruction from a single image is a challenging problem, especially under partial occlusions and extreme poses. This is because the uncertainty of the estimated 2D landmarks will affect the quality of face reconstruction. In this paper, we propose a novel joint 2D and 3D optimization method to adaptively reconstruct 3D face shapes from a single image, which combines the depths of 3D landmarks to solve the uncertain detections of invisible landmarks. The strategy of our method involves two aspects: a coarse-to-fine pose estimation using both 2D and 3D landmarks, and an adaptive 2D and 3D re-weighting based on the refined pose parameters to recover accurate 3D faces. Experimental results on multiple datasets demonstrate that our method can generate high-quality reconstruction from a single color image and is robust for self-occlusions and large poses.