Multi-objective optimization for X-ray exposure reduction during image-guided needle-based procedure

Nowadays, an increasing number of surgical operations are guided by intra-operative imaging, especially Cone Beam Computed Tomography (CBCT) systems using X-rays. We deal with the problem of reducing the X-ray dose received by the patient and the staff while ensuring reasonable visibility of the region of interest during needle-based ablation procedures. We achieved this by implementing a numerical multi-optimisation algorithm using as input numerical X-ray simulation providing both medical staff and patient exposure, as well as an estimation of the fluorescence-based ablation needle visibility. Our algorithm can find the Pareto front in this multi-objective function problem. The Pareto front is suggested to the operator who can then select the best configuration according to the defined priority level regarding each exposure and visibility. We have been able to reduce in simulations the clinician's dose by 61.8%, the patient's dose by 45.1% and increase the region of interest visibility by 28.6% in-plane axis view and, respectively, by 41.5%, 60.4% and 41.0% in progression view.