Click here to


Are you sure ?

Yes, do it No, cancel

Development of a Patient Specific Collision Detection Algorithm for Non-Coplanar Cone Beam CT in Stereotactic Radiosurgery Using Eclipse Scripting

T Mann1,2*, N Ploquin1,2, T Novak2, K Thind1,2, (1) University of Calgary, Calgary, AB, (2) Tom Baker Cancer Center, Calgary, AB


(Tuesday, 7/31/2018) 3:45 PM - 4:15 PM

Room: Exhibit Hall | Forum 5

Purpose: The main obstacle in acquiring non-coplanar CBCT clinically is not being able to predict whether a given couch angle will allow the OBI to perform a CBCT without collision. Through the use of eclipse scripting this project seeks to check, via automation in Eclipse, all patient specific potential collisions between the gantry, OBI system, couch, and the patient at the treatment planning stage.

Methods: ESAPI (Eclipse Scripting Application Programming Interface) by Varian provides access to the Eclipse treatment plan data. Our collision detection model extracts patient isocenter location, body contour, couch location, and treatment beam data from Eclipse using ESAPI. Contour data is saved as geometric meshes and correlated with a basic model of the gantry head and OBI system using the plan isocenter location. The gantry and OBI are rotated around the patient while our simulation performs a point to point analysis between structures to determine if the minimum distance is reached. Using a safety margin, a collision was defined as structures being closer than 5 cm. To test the validity of our simulation four different plans with separate isocenter locations and multiple beams, representing limiting cases with potential collision in SRS, were created for an anthropomorphic body phantom. Minimum distances and the gantry angle at which they occurred were found for the four isocenters; left and right, occipital and frontal, with various couch angles.

Results: The average total difference between simulated and measured results for 56 beams was found to be 1.4 cm with a standard deviation of 1.8 cm and a 100% collision detection rate using a 5 cm margin.

Conclusion: This current work is a viable solution for automated SRS collision modelling with an acceptable degree of accuracy. Future work will include implementation of a faster collision testing algorithm such as bounding sphere hierarchies.


Modeling, Stereotactic Radiosurgery, Cone-beam CT


IM/TH- RT X-ray Imaging: CBCT imaging/therapy implementation

Contact Email