Radiofrequency ablation (RFA) therapy is recognized as a safe and effective method of treating certain types of tumor, particularly those in the liver and lung. The therapy involves the image-guided percutaneous insertion of an RFA needle electrode into the tumor. Radiofrequency current delivered to the electrode creates ionic agitation and tissue heating around the needle tip, which causes cell death by coagulation necrosis above 50 degrees C.
The procedure aims to kill the entire tumor plus a small surgical margin of surrounding tissue. Its success depends on the clinician’s skill in localizing the tumor and accurately covering the tumor with multiple ablations by placing one or more RFA needles, typically using Computed Tomography (CT) and/or ultrasound image-guidance.
The success rate of RFA therapy is currently greatest for tumors with a maximum diameter of about 4 cm, this being the largest volume that can be treated with a single RFA needle insertion. Larger tumors require multiple overlapping RFAs, making the procedure much more difficult to plan and execute. There is therefore a clinical need for tools that assist clinicians with therapy planning and needle placement, and that provide real-time feedback on tumor ablation so that they can adapt therapy during the course of the procedure.
RFA cockpit concept
To address the above need, Philips Research is exploring a comprehensive RFA cockpit concept for the treatment of liver cancer that integrates RFA therapy planning, navigation and feedback.
The Philips RFA cockpit concept is essentially a coherent user interface designed around the clinician. For therapy planning, the cockpit leverages the tumor visualization capabilities of pre-operative CT and/or PET imaging plus automated generation of target ablation plans. For targeted needle placement, it combines these with the real-time imaging capabilities of ultrasound and precision needle-tip navigation technology that is based on electromagnetic tracking.
Electromagnetic tracking involves the use of miniaturized sensor coils that are built into interventional instruments. By processing signals that the sensor coils pick up from an applied electromagnetic field, the position of the RFA needle-tip can be localized in space during the interventional procedure.
The real-time spatial data that is acquired allows accurate fusion of real-time imaging and needle visualization with the therapy planning images. Intra-procedural imaging also provides rapid feedback on tumor ablation, which can be used to iteratively adapt the therapy during the procedure and thereby help to achieve full tumor coverage.
In order to assess the technical feasibility and usability of the RFA cockpit concept, Philips Research has developed an experimental system that is currently being evaluated in clinical studies at the National Institutes of Health Clinical Center (Bethesda, MD, USA).
Accurate therapy planning requires target tumor identification and delineation using pre-operative images. The experimental system therefore includes multi-modal imaging tools to fuse pre-operative CT and/or PET images with real-time ultrasound images, plus advanced segmentation software for tumor delineation.
Based on the 3-dimensional tumor segmentation, the estimated ablation volumes and the preferred orientation for inserting the RFA needles, the therapy planning algorithms compute an RFA plan. This plan consists of the computed locations of the optimal number of ablations to treat the entire tumor volume, and visualizes the tumor segmentation, the multiple overlapping ablations and the composite treated volume.
During the RFA procedure, real-time 3-dimensional navigation data provided by the needle-tip navigation system guides placement of the RFA needles to the precise locations as defined by the computed RFA plan. Simultaneous real-time ultrasound imaging continuously keeps the segmented tumor model (including the target locations) accurately positioned within the navigation space. As a result, any deviation between the planned needle positions and those actually achieved becomes immediately apparent.
Real-time ultrasound imaging and/or CT scans performed periodically during the procedure provide information on the volume of ablated tissue, which is used to iteratively compute an updated ablation plan for residual tumor ablation. The updated plan is used to navigate subsequent needle placements, or to introduce additional ablations with the objective of achieving complete tumor coverage. The computation time required to generate these plans is less than 30 seconds. During system evaluation studies, plans were developed and adapted quickly either before or during the ablation procedure.
Evaluated in more than 60 patients, the experimental system achieved a navigation accuracy of 3.8 mm. This is within the clinical requirements for successful execution of most RFA procedures. For ablation targets in critical locations that require even better accuracy, Philips Research has already developed advanced registration methods that improve the navigation accuracy to 2.9 mm. It has also demonstrated the real-time fusion of pre-operative PET scans into the therapy planning stage in order to target tumors that are not easily visualized using CT or ultrasound imaging.