Evaluation of a 4D Cone-Beam CT Reconstruction Approach Using an Anthropomorphic Phantom

We have previously developed image-guided navigation systems for thoracic abdominal interventions utilizing a three dimensional (3D) Cone-Beam CT (CBCT) image acquired at breath-hold. These systems required the physician to perform the intervention in a gated manner, with actions performed at the same respiratory phase in which the CBCT image was acquired. This approach is not always applicable, as many patients find it hard to comply with the breath-hold requirement. In addition the physician’s actions are limited to a specific respiratory phase. To mitigate these deficiencies we have developed and implemented a retrospectively gated acquisition protocol using a clinical C-arm based system. The resulting 4D (3D+time) image is then used as input for the navigation system. We evaluate our reconstruction approach using a computer controlled anthropomorphic respiring phantom. The phantom is respired using respiratory rates of 12, 15 and 20 breaths per minute, and three amplitudes corresponding to shallow, normal, and deep breathing patterns. We show that the gated images have a better contrast to noise ratio and sharper edges than the images reconstructed without gating. Thus we are able to acquire an intra-operative data set that potentially provides better navigation accuracy, using 3D images at arbitrary points in the respiratory cycle without requiring the patient to hold their breath during image acquisition.