Registration of 3-D heart volumes to 2-D fluoroscopy images

Cardiac ablation is a treatment option for atrial fibrillation which is the most common heart arrhythmia. It is done using catheters, usually under fluoroscopic guidance. While the catheters can be clearly seen in the fluoroscopic images, the heart remains invisible in the X-ray images. However, knowing the position of the catheters inside the heart is crucial. To highlight the heart structure, the physician injects contrast agent. Also an overlay showing the patient's heart can be displayed together with the fluoroscopic images.

Registration of the 3-D heart volume to 2-D fluoroscopic images is needed as the patients position during the acquisition of the heart model and the position during the intervention differ. To avoid time consuming manual registration we develop automatic approaches to perform this registration.

Structure-Enhancing Visualization for Manual Registration in Fluoroscopy

Electrophysiology procedures are in many cases performed under fluoroscopic guidance. As the heart is not visible under fluoroscopy, a preoperatively acquired heart model of the patient may be fused with the live X-ray images to provide the physician with a better orientation. This heart model needs to be registered to the fluoroscopic images. Currently, registration is performed manually, e.g., after contrast injection into the left atrium during atrial fibrillation procedures. We propose a novel visualization of the heart model that shows the same structures as the contrast agent and allows an easy identification of corresponding landmarks and features in both the model and the angiography images.

View Poster

Link to paper coming soon.

Contact

Address

Dr.-Ing. Matthias Weidler

Alumnus of the Pattern Recognition Lab of the Friedrich-Alexander-Universität Erlangen-Nürnberg

Registration of 3-D heart volumes to 2-D fluoroscopy images



	Website deprecated and outdated. Click here for the new site.

Department of Computer Science 5 Our Team Hoffmann, Matthias Projects Registration of 3-D heart volumes to 2-D fluoroscopy images Catheter Detection and Reconstruction Lectures Publications Thesis Curriculum Vitae Research Publications Free Software Data Courses Curriculum Theses Press Releases Cooperations Open Positions LME Videos Ph.D. Gallery Contact Intranet Impressum Datenschutzerklärung Contact Address Universität Erlangen-Nürnberg Chair of Computer Science 5 (Pattern Recognition) Germany Driving directions Powered by	Dept. of Computer Sc. » Pattern Recognition » Our Team » Hoffmann, Matthias » Projects » Registration of 3-D heart volumes to 2-D fluoroscopy images Dr.-Ing. Matthias Weidler Alumnus of the Pattern Recognition Lab of the Friedrich-Alexander-Universität Erlangen-Nürnberg Registration of 3-D heart volumes to 2-D fluoroscopy images Cardiac ablation is a treatment option for atrial fibrillation which is the most common heart arrhythmia. It is done using catheters, usually under fluoroscopic guidance. While the catheters can be clearly seen in the fluoroscopic images, the heart remains invisible in the X-ray images. However, knowing the position of the catheters inside the heart is crucial. To highlight the heart structure, the physician injects contrast agent. Also an overlay showing the patient's heart can be displayed together with the fluoroscopic images. Registration of the 3-D heart volume to 2-D fluoroscopic images is needed as the patients position during the acquisition of the heart model and the position during the intervention differ. To avoid time consuming manual registration we develop automatic approaches to perform this registration. Structure-Enhancing Visualization for Manual Registration in Fluoroscopy Electrophysiology procedures are in many cases performed under fluoroscopic guidance. As the heart is not visible under fluoroscopy, a preoperatively acquired heart model of the patient may be fused with the live X-ray images to provide the physician with a better orientation. This heart model needs to be registered to the fluoroscopic images. Currently, registration is performed manually, e.g., after contrast injection into the left atrium during atrial fibrillation procedures. We propose a novel visualization of the heart model that shows the same structures as the contrast agent and allows an easy identification of corresponding landmarks and features in both the model and the angiography images. View Poster Link to paper coming soon.