Anatomical Modeling to Improve the Precision of Image Guided Liver Ablation

点击翻译按钮获取中文摘要

Primary and secondary liver cancers are increasing in incidence and are collectively responsible for over 1 million deaths per year worldwide. Among the curative treatments available for liver cancers, surgical resection is considered the standard of care. Unfortunately, less than 20% of patients are eligible for such resection at the time of the diagnosis. Image-guided percutaneous thermal ablation (PTA) has become a widely utilized option for patients not eligible for surgery with local control success rates ranging from 55% to 85% (4-6). In order to achieve optimal results following PTA, rates of residual tumor or recurrence should be minimized (6, 8), which can be achieved by providing adequate minimal ablation margins around the tumor. To meet this goal, it is critical to have high-quality intra-procedurally imaging that offers information in respect precise definition of extent of the target tumor, confirmation of ablation probe placement at the target tumor(s), and accurate ablation margins assessment. Currently, there are no commercially available tools that enable an accurate method for tumor mapping and ablation assessment while taking in consideration biomechanical conformational changes associated with the ablation therapy. Based in our preliminary work, we hypothesize that local tumor control following ablation of liver cancers will be improved with the application of a dedicated anatomical linear elastic biomechanical model for treatment guidance and efficacy assessment by enabling accurate identification and targeting of the tumor and providing intra-procedural assessment of the ablation, respectively. This hypothesis will be tested through three specific aims. Firstly, we will optimize the anatomical modeling liver ablation guidance in the RayStation Platform by validating the accuracy of the linear elastic biomechanical models of the liver for the application of mapping the tumor defined on the pre-interventional images onto the intra-procedural images obtained just prior to ablation; Secondly, we will evaluate the impact of this model on local tumor control following liver ablation by conducting a phase II randomized clinical trial; Finally, we will optimize the biomechanical model to enable modeling of the local changes in the tumor and surrounding normal tissue resulting from the ablation. We believe that the integration of accurate, precise, and efficient biomechanical modeling tools to determine the tumor location at the time of ablation and to monitor the ablation margin will improve local tumor control rates in patients with liver cancers, potentially improving overall survival rates. The ability to perform deformable image registration to map the tumor, identified on pre-intervention imaging, in the presence of artifacts from the ablation probe and with little to no contrast within the liver presents a significant challenge to most intensity-based algorithms. The use of a biomechanical-based model in this application is poised to make a significant impact, potentially enabling local control for the 20% of patients who fail this therapy. The integration of this technology into the RayStation platform ensures that this technology is widely available to patients.