Subscribe to RSS
DOI: 10.1055/s-0036-1584048
A web-based and user friendly planning tool to predict outcome for percutaneous microwave ablation of liver tumors including validation of the simulation model
Aims and objectives: The aim of our study was to develop and validate a web-based planning tool to predict outcome for percutaneous image-guided microwave ablation for liver tumors.
Methods and materials: A web-based environment including a viewer and image segmentation and registration tools was developed. A simulation model was incorporated. Pseudo-anonymized scans of microwave procedures were used to validate the performance of the developed simulation model.
First, the pre-interventional CT was uploaded and the liver and tumor were semi-automatically segmented. Next, intra-operative scans were uploaded and registered to the pre-operative images. This generated needle coordinates likewise to the original situation. After placing virtual needles using these coordinates, equipment parameters and treatment protocols were set. A simulation was executed which estimated the microwave lesion. The physical (real) lesion was segmented using a 1-month post-microwave follow-up scan and quantitatively compared to the predicted lesion. Outcome measures of this comparison were the average absolute error (α) between the segmented (S) and simulated (Σ) lesion and the volumetric ratio called target overlap (φs). Underestimation of the simulated lesion is shown by φs < 1.0. Overestimation, where S < Σ, is indicated by large α and (φs) ˜ 1.
Results: Thirty five cases were successfully tested and up to now five were used for validation. Mean (range) for φs and α (in mm) were 0,9414 ± 0,2542 and 2,9119 ± 1,8772 respectively.
Conclusion: We present a planning tool to predict outcome for percutaneous microwave ablation of liver tumors. Based on first validation results, further refinement of the simulation model is needed.