Title of article :
Evaluation of robot-based registration for subtraction radiography
Author/Authors :
Grigore C. Burdea، نويسنده , , Stanley M. Dunn، نويسنده , , Guy Levy، نويسنده ,
Issue Information :
روزنامه با شماره پیاپی سال 1999
Abstract :
Digital subtraction of (plane film) radiographs of an area of interest taken over time is a powerful diagnostic tool. In dentistry for example, this is used to detect and treat periodontal disease (i.e. the sequence of diseases that leads to the loss of bone supporting the teeth). A precondition of subtraction radiography is correct three-dimensional registration of the film and X-ray source, without which the subtraction images are meaningless. A small misalignment can be compensated for by post-processing of the digitized films (by correspondence and rectification) to remove artifacts due to positioning errors. This paper presents a novel approach to subtraction radiography which replaces customary mechanical alignment methods (such as a stent or cephalostat) with a robot. A mechanical stent is a short rod attached at one end to the X-ray source and at the other to a mechanical appliance protruding from the patientʹs mouth. An experimental system was constructed which creates a ‘sensorized stent’ by integrating a plastic mouth appliance (with the impression of the patientʹs teeth), a robot carrying an X-ray source and a host computer. Results showed that the robotic system was superior to the mechanical alignment approach, due to its excellent accuracy and repeatability. This resulted in much less variation in the non-registered X-ray images, and in a smaller standard deviation in the intensities of subtracted images overall. The results suggest that in the future, diagnostic studies including subtraction radiography will not need either mechanical alignment (which is imprecise) or post-processing registration (which is time consuming).
Keywords :
Dental radiograph , image registration , Robot , subtraction radiography , tracking
Journal title :
Medical Image Analysis
Journal title :
Medical Image Analysis