Experimental measurement of human head motion for clinical dental CBCT system design

Theoretical speaking, CT technique reconstructs cross-sectional images of an object from a set of projections taken from many different angle views. In the commercial CT, these projections are continually acquired over a certain period of time with an assumption that the object being scanned is stationary. However, the motionless assumption is unpractical in clinical applications because some motions of human organs are unavoidable, such as peristalsis, heart beating and other physiological motilities. These motions may result in motion artifacts such as blurring or doubling, which are well-known in medical imaging filed including not only CT but also other imaging modalities like MRI, PET and SPECT. In recent years, some new kinds of cone-beam CT (CBCT) have been produced by using large area flat-panel detectors which have very high spatial resolutions. A typical example is dental CBCT system. Since the human organ motions are unavoidable, the evaluation of motion effects should be done during the design of a new clinical CT system. In this paper, we designed and completed an experiment on 37 volunteers to simulate a typical acquisition of commercially dental CBCT where the patient usually sits down. Our goal is to identify the minimal movements of the human head under ideal conditions in the attitudes of sitting. All the volunteers are healthy and tractable who were asked to keep calm without pressures. This paper presents our devices and methods for head motion measurements which are based on the binocular stereo vision technique. Then, the experimental results and analysis are presented. In the end, the effects of these head motions are discussed on a new clinical high-resolution CT system design.