Publication details

Accuracy of registration for fusion imaging with ultrasound navigation system


ANDRAŠINA Tomáš KALA Ondřej FOUKAL Jakub VÁLEK Vlastimil

Year of publication 2014
Type Appeared in Conference without Proceedings
MU Faculty or unit

Faculty of Medicine

Description Purpose The Philips iU22 PercuNav 4.0 (Koninklijke Philips Electronics N.V.) is an image navigation system, which provides real-time, three-dimensional visualization, and navigation tools for all stages of diagnosis and intervention. In this study, we analyze the accuracy of different registration methods for fusion imaging using the Percunav system on a phantom model. Material and Methods A preprocedural CT dataset was obtained using interventional triple modality phantom (Fluke Biomedical). Three different methods were used for the registration of real-time ultrasound images with CT datasets (registration patch match, ultrasound plane match, and US point match). For each method, variable distance of matching points to the target lesion was evaluated. Each subgroup fusion comprised 12 attempts with two operators. The real deviation of the target lesion depicted on CT scan and ultrasound in an anterior approach (approach for fusion), and the lateral approach were measured. Student’s t test and F test were used for statistical analyses. Results The real accuracy of CT fusion and ultrasound is not linearly dependent on the overall fit, counted by navigation systems after 3D data registration. US plane match registration is less time-consuming compared with 3 points match; it shows a good correlation between the overall fit and the target lesion in the anterior approach for distant lesions in the anterior and lateral approaches. There is a statistically significant difference in accuracy for depicting the target lesion from the anterior and lateral approaches (p<0.05). Conclusion All registration methods have comparable precision for targeting lesions in fusion imaging. Using different approaches for fusion registration and navigation could lead to significant inaccuracy of systems.