Effects of motion blurring in x-ray fluoroscopy.

In conventional fluoroscopy, continuous x-ray exposure blurs moving objects, while in pulsed fluoroscopy, short duration x-ray pulses acquire images without motion blur. Many perception experiments on noisy image sequences are consistent with low-pass temporal filtering by the human visual system, and this is anticipated to cause visual system "blurring" of moving objects. With moving cylinders in spatially white noise, we simulated 30 acq/s (acquisitions per second), continuous fluoroscopy having both x-ray and visual system motion blur. We also simulated pulsed fluoroscopy at 30 acq/s (pulsed-30) having visual system but not x-ray system motion blur. For both continuous and pulsed-30 acquisitions, with increasing velocity, detectability of small cylinders decreased by as much as approximately 50%, while detectability of large cylinders increased and then decreased. Detectability of pulsed-30 was only slightly higher than continuous, indicating that visual system motion blurring dominated x-ray system blurring. For the case of stationary objects, blurring greatly reduced detectability, indicating that last-image-hold of moving objects deteriorates with continuous acquisitions. With no free parameters, a human observer model with an independently measured spatio-temporal contrast sensitivity function accurately described all effects.