University of Twente Student Theses


Tracked ultrasound : towards accurate intraoperative navigation during pediatric bone tumor resections with soft tissue components

Zee, J.M. van der (2023) Tracked ultrasound : towards accurate intraoperative navigation during pediatric bone tumor resections with soft tissue components.

[img] PDF
Abstract:Purpose: Resection of pediatric osteosarcoma in the extremities with soft tissue involvement presents surgical challenges due to difficult visualization and palpation of the tumor. Therefore, an adequate image-guided surgery (IGS) system is required for accurate tumor resection. Ideally, an IGS system should be combined with a fast registration and re-registration methodology that minimal interrupts the surgical workflow. Moreover, the surgical decision-making may be enhanced with the use of a three- dimensional patient-specific model in combination with intraoperative tracked ultrasound (iUS). This study determines the surgical feasibility of an iUS-based IGS tool that incorporates the bone surface found with iUS as a new registration feature. Methods: An automatic bone segmentation pipeline (Part I) and a surgical navigation system (Part II) were developed. Ultimately, the derived system was evaluated by pediatric surgeons (Part III). In Part II, a 3D model of the porcine lower limb was created based on preoperative scans. Second, the bone surface of the tibia was automatically detected (Part I) on an iUS acquisition. This acquisition was performed with a free-hand sweep over the skin. The bone surface of the preoperative 3D model was then matched with the bone surface detected by the iUS. To evaluate the registration accuracy, ten artificial surgical targets were used to calculate the target registration error (TRE). In Part III, the intraoperative performance of this surgical tool was evaluated by six pediatric surgeons and two pediatric oncologic orthopedists. The participants had to localize ten surgical targets to compute the point localization error (PLE). Finally, the user-experience was assessed with a post-procedural questionnaire. Results: An automatic segmentation algorithm for bone segmentation was developed. Within the developed surgical system, a mean TRE of 6.78 mm (STD = 0.70 mm) and a median PLE of 8.00 mm (3.86 – 8.96 mm). The participated surgeons were more confident about their actions, and they experienced the localization as faster and easier concluding from the answers in the user-experience questionnaire. Conclusion: This study indicates the potential and concept of an iUS-based registration methodology for an IGS system in a porcine cadaver study. Moreover, the participated surgeons and orthopedists were convinced of the clinical value of the interaction between the iUS and the 3D model
Item Type:Essay (Master)
Faculty:TNW: Science and Technology
Programme:Technical Medicine MSc (60033)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page