Comparison of intraoperative MR imaging and 3D-navigated ultrasonography in the detection and resection control of lesions.

Abstract

OBJECT: The authors undertook a study to compare two intraoperative imaging modalities, low-field magnetic resonance (MR) imaging and a prototype of a three-dimensional (3D)-navigated ultrasonography in terms of imaging quality in lesion detection and intraoperative resection control. METHODS: Low-field MR imaging was used for intraoperative resection control and update of navigational data in 101 patients with supratentorial gliomas. Thirty-five patients with different lesions underwent surgery in which the prototype of a 3D-navigated ultrasonography system was used. A prospective comparative study of both intraoperative imaging modalities was initiated with the first seven cases presented here. In 35 patients (70%) in whom ultrasonography was performed, accurate tumor delineation was demonstrated prior to tumor resection. In the remaining 30% comparison of preoperative MR imaging data and ultrasonography data allowed sufficient anatomical localization to be achieved. Detection of metastases and high-grade gliomas and intraoperative delineation of tumor remnants were comparable between both imaging modalities. In one case of a low-grade glioma better visibility was achieved with ultrasonography. However, intraoperative findings after resection were still difficult to interpret with ultrasonography alone most likely due to the beginning of a learning curve. CONCLUSIONS: Based on these preliminary results, intraoperative MR imaging remains superior to intraoperative ultrasonography in terms of resection control in glioma surgery. Nevertheless, the different features (different planes of slices, any-plane slicing, and creation of a 3D volume and matching of images) of this new ultrasonography system make this tool a very attractive alternative. The intended study of both imaging modalities will hopefully allow a comparison regarding sensitivity and specificity of intraoperative tumor remnant detection, as well as cost effectiveness.