Physiological motion modeling for organ-mounted robots

Background: Organ-mounted robots passively compensate heartbeat and respiratory motion. In model-guided procedures, this motion can be a significant source of information that can be used to aid in localization or to add dynamic information to static preoperative maps. Methods: Models for estimating periodic motion are proposed for both position and orientation. These models are then tested on animal data and optimal orders are identified. Finally, methods for online identification are demonstrated. Results: Models using exponential coordinates and Euler-angle parameterizations are as accurate as models using quaternion representations, yet require a quarter fewer parameters. Models which incorporate more than four cardiac or three respiration harmonics are no more accurate. Finally, online methods estimate model parameters as accurately as offline methods within three respiration cycles. Conclusions: These methods provide a complete framework for accurately modelling the periodic deformation of points anywhere on the surface of the heart in a closed chest.