2019 IEEE Nuclear Science Symposium and Medical Imaging Conference (NSS/MIC) | 2019

Motion-Adaptive Gantry Development for Open-Field Mouse PET

 
 
 
 
 

Abstract


How the mammalian brain responds to changes in the environment is still poorly understood. We are developing the Open-Field positron emission tomography (PET) system to study brain-behavior-environment interactions in freely moving mice. This system overcomes an important technological limitation of existing small animal PET systems by enabling motion-compensated PET imaging of the fully conscious mouse brain concurrently with analysis of an animal’s behavioral response during cognitive and/or pharmacological challenges. A key component of the system is the motion-adaptive gantry which allows the tomograph to slide linearly in response to animal motion. Here we report the gantry design and initial closed-loop motion control performance of the gantry for a realistic load and simplified motion profiles.The gantry supports two rails of length 1.5 m, along which the main carriage translates via a belt-drive and brushless servo motor. Motor control involves a Galil sine drive motion controller which implements a proportional-derivative-integral (PID) controller and relies on feedback from the built-in motor encoder. The controller response to linear and sinusoidal reference motion profiles was tested using both default and manually optimized PID control parameters, Kd, Kp, Ki, and the feed-forward velocity (FV), acceleration (FA) and integrator limit (IL) parameters.Control accuracy and synchronization was poor for the sinusoidal profiling at a frequency of 0.5 Hz with 10000 counts (0.5 m) amplitude, exhibiting obvious overshoot and lag. Following manual tuning (Kd = 130, Kp = 20, Ki = 0, IL = 9.982V, FV = 15, FA = 0) the system exhibited minimal overshoot with <1% error in the velocity profile for a step motion and <5% error for a sine wave of 0.5 Hz and 10000 counts amplitude.The next steps of this work will focus on achieving optimized control for real mouse motion profiles and integrating motion tracking to provide reference motion profiles for the controller.

Volume None
Pages 1-3
DOI 10.1109/NSS/MIC42101.2019.9059887
Language English
Journal 2019 IEEE Nuclear Science Symposium and Medical Imaging Conference (NSS/MIC)

Full Text