OBJECT: During routine quality assurance (QA) testing of the gamma knife Automatic Positioning System (APS), it was noticed that slight pressure applied to the handle of the QA test tool produced visible rotation of the APS assembly. The authors describe the tests performed to quantify these movements. METHODS: A stereotactic frame was positioned in the slides of the APS and known loads applied using a spring balance to the upper corner of the frame. Movement at the corner of the frame (y = 200) was measured using a dial gauge. The typical load applied to the APS by a patient with a medium-sized head was determined in a treatment simulation. Projecting the measured data, a plot showing the resultant error in the position of a target point was constructed. Error values increased with distance from the frame, up to a maximum of 0.3 mm (at the superior limit of treatment range). Increased loads could be applied to the system during patient movements, and these would result in larger displacements. Two volunteers enlisted to simulate patient movements and the deflections were recorded. It was estimated that maximum errors in the target positioning of 0.45 mm might take place during these movements. CONCLUSIONS: The use of the APS has some additional and unexpected associated errors. The authors believe, however, that because the errors are smaller than those of target localization and manual trunnion treatments, its clinical use is still justified. The manufacturer is investigating the cause of the movements and possible solutions.
OBJECT: During routine quality assurance (QA) testing of the gamma knife Automatic Positioning System (APS), it was noticed that slight pressure applied to the handle of the QA test tool produced visible rotation of the APS assembly. The authors describe the tests performed to quantify these movements. METHODS: A stereotactic frame was positioned in the slides of the APS and known loads applied using a spring balance to the upper corner of the frame. Movement at the corner of the frame (y = 200) was measured using a dial gauge. The typical load applied to the APS by a patient with a medium-sized head was determined in a treatment simulation. Projecting the measured data, a plot showing the resultant error in the position of a target point was constructed. Error values increased with distance from the frame, up to a maximum of 0.3 mm (at the superior limit of treatment range). Increased loads could be applied to the system during patient movements, and these would result in larger displacements. Two volunteers enlisted to simulate patient movements and the deflections were recorded. It was estimated that maximum errors in the target positioning of 0.45 mm might take place during these movements. CONCLUSIONS: The use of the APS has some additional and unexpected associated errors. The authors believe, however, that because the errors are smaller than those of target localization and manual trunnion treatments, its clinical use is still justified. The manufacturer is investigating the cause of the movements and possible solutions.