Home
Annual Meeting
Winter Workshop
Committees
Join ISMICS
Members Only
Member Search
Journal
Newsletters


Concept of Active Motion Compensation: Modeling and prediction of aortic root motion to guide automatically implantation of suture less stent-fixed aortic valve prosthesis

Back to Annual Meeting
Back to Program


Stephan Jacobs, Volkmar Falk.
Cardio-Thoracic-Surgery, Heartcenter Leipzig, Leipzig, Germany.

Objective: Recently developed techniques open perspectives of heartbeat synchronized navigation and positioning of artificial aortic valve replacements. Using a new concept of active motion compensation, modeling and prediction of aortic root motion provides guidance for automatically implantation of suture less transcatheter guided stent-fixed aortic valves.
Methods:
With heart beat synchronization the motion between the surgical tool and the aortic root was eliminated to reduce the complexity of the surgical task. The synchronization contains a measuring system to detect the motion, an algorithm to predict the motion and to compute the inputs of the robot control and a system to position the aortic valve replacement. Based on two data sets (ultrasound and CT image data), an algorithm for segmentation and reference point detection (natural landmarks) was developed. Subsequently image data sets were segmented in appropriate way to extract geometric shape of the aortic root. To predict the trajectories of the reference point, adaptive filter were designed and subsequently compared.
Results:
An algorithm for automatic segmentation of the aortic root was developed, which allowed visualisation of the aortic root at different times by surface models. For the target position of the prosthesis within the aortic root, a correlation between the target position of the prosthesis and some natural landmarks (commissures) of the aortic root could be found. An algorithm was developed to detect natural landmarks (commissures) within the ultrasound images of the aortic root.
To compensate the delays of the complete process of heart beat synchronization, a prediction of the motion of the natural landmarks based on adaptive filtering is proposed. Using the predicted positions of the landmarks and the functional reference, the target position of the prosthesis can be calculated in advance and can reach this position in time.
Conclusion:
This method for modeling and prediction of the aortic root motion provides guidance for automatically percutaneous or transapical aortic valve replacement. First trials showed promising results and have to be determined in further studies


Back to Annual Meeting
Back to Program

About Us | Contact Us | Privacy Policy
Copyright© The International Society for Minimally Invasive Cardiothoracic Surgery. All Rights Reserved.