MRI based assessment of the extent to which stereo-tracking of markers on the chest can predict motion of the heart

摘要: We have developed a visual-tracking-system (VTS) which uses stereo-imaging to track the motion of markers on patients during cardiac SPECT imaging with goal using tracked correct for patient motion. The aim this study is determine MRI in volunteers if rigid-body-motion (RBM) model can be used predict heart within chest from surface chest. Our methodology investigating body-motion separate influence respiration volunteer hold their breath acquisition sequence 2 sets EKG-triggered sagittal slices covering heart, first set pre-motion and second post-motion. link these studies such that process takes ∼ 45 seconds initiation till given instruction resume breathing. An analysis combined individual obtain an estimate six-degree-of-freedom (6DOF) RBM volunteer. estimated by semi-automatic 3D segmentation region subsequent registration slices. Studies did not intentionally move between acquisitions were establish baseline agreement methods estimation small motions. VTS estimates was good “no-motion” case maximum difference ≪ 5 mm any translation component, ≪2 deg rotational component. With body translations as axial-slide might expected approximate RBM, also being 6 mm, 2.5 deg. Application non-RBM counter-clock-wise (CCW) twist at shoulders hips fixed, decreased magnitude misalignment caused but same extent previous classes differences class motions ≪11 4.5 angular. Thus use predictions correction reconstruction should decrease artifacts types studied, less so those are better modeled non-RBM.