Image Fusion for Computer-Assisted Bone Tumor Surgery

Abstract

Conventionally, orthopaedic tumor surgeons have to mentally integrate all preoperative images and formulate a surgical plan. This preoperative planning is particularly difficult in pelvic or sacral tumors due to complex anatomy and nearby vital neurovascular structures. At the surgery, the implementation of these bone resections are more demanding if the resections not only are clear of the tumor but also match with custom implants or allograft for bony reconstruction. CT and MRI are both essential preoperative imaging studies before complex bone tumor surgery. CT shows good bony anatomy, whereas MRI is better at indicating tumor extent and surrounding soft tissue details. Overlaying MRI over CT images with the same spatial coordinates generates fusion images and 3D models that provide the characteristics of each imaging modality and the new dimensions for planning of bone resections. This accurate planning may then be reproduced when it is executed with computer-assisted surgery. It may offer clinical benefits. Although current navigation systems can integrate all the preoperative images for resection planning, they do not support the advanced surgical planning that medical engineering CAD software can provide, such as virtual bone resections and assessment of the resection defects due to system incompatibility. Translating the virtual surgical planning to computer navigation by manual measurements may be prone to processing errors. Currently, the integration of CAD data into navigation system is made possible by converting the virtual plan in CAD format into navigation acceptable DICOM format. The fusion of the modified image datasets that contain the virtual planning with the original image datasets allow easy incorporation of virtual planning for navigation execution in bone tumor surgery. Image fusion technique has also been used in bone allograft selection from a 3D virtual bone bank for reconstruction of a massive bone defect. This facilitates and expedites in finding a suitable allograft that matches with the skeletal defect after bone tumor resection. The technique is also utilized in image-to-patient registration of preoperative CT/MR images. It takes the forms of 2D/3D or 3D/3D image registration. It eliminates the step of surface registration of the operated bones that normally require large surgical exposure. The registration method has great potential and has been applied for minimally invasive surgery in benign bone tumors. This article provides an up-to-date review of the recent developments and technical features in image fusion for computer-assisted tumor surgery (CATS), its current status in clinical practice, and future directions in its development.