Skip to Main Content
Patient motion, which causes artifacts in reconstructed images, can be a serious problem in SPECT imaging. If patient motion can be detected and quantified, the reconstruction algorithm can compensate for the motion. Our approach is based on optical tracking of the patient using a pair of Web cameras to acquire stereo images. The cameras, mounted outside a SPECT system, acquire optical images simultaneously with the emission projections. The patient wears a stretchable, close-fitting garment that includes easily detected features. The Web cameras view the features, from which a surface map is computed using stereo techniques. When the patient moves, the surface map is recomputed, the patient surface is tracked, and a description of the motion is computed. Later processing stages can use the motion information to correct the tomographic reconstruction, for example, by rebinning acquired projection data. In this investigation, we examined whether patient respiratory motion can be detected using this approach. We have found that features can be reliably tracked over time. Feature motion as computed from stereo measurements was compared with patient respiration measured independently using a pneumatic bellows. The two sources of motion information were found to be highly correlated, demonstrating that stereo optical imaging can detect patient motion. Lastly, we have been able to compute translational and rotational motion parameters to describe respiratory motion.