There is growing evidence that maximal surgical resection of malignant gliomas is beneficial, both by increasing the progression free survival (Stummer, Pichlmeier et al. 2006) and also by facilitating postoperative chemotherapy and radiotherapy (Stupp, Mason et al. 2005). In this context there has been an increased need for real time intraoperative imaging in brain tumour surgery. The complex theatre arrangements, prohibitive cost and prolonged theatre time has restricted the wide application of intraoperative MRI (iMRI). Modern intraoperative ultrasound, which in the past has been relatively underused in oncological neurosurgery, is affordable, and easy to use (Unsgaard, Ommedal et al. 2002), but has not been robustly validated.
This study attempts to histologically validate the accuracy of the intraoperative US and to explore its potential as an intraoperative navigation tool, which can accurately guide biopsies and resections of intrinsic brain tumours. The digital data extracted from US images obtained during ultrasound guided biopsies was correlated with the histology of the relevant specimens. Image analysis of selected regions of interest (ROI) was employed to extract quantitative parameters from the digital ultrasound images. The mean pixel brightness (MPB) and the standard deviation (SD) were correlated with histological parameters. The pattern of histograms from the selected ROIs was observed and correlated with the histological findings. A close correlation was observed between mean pixel brightness (MPB), an objective measure of echogenicity, and cellularity and an equally close correlation between the standard deviation (SD) and the intrinsic cellular diversity of the analysed areas. These two together, despite not being specific, can indirectly suggest the nature of the tumour and also reflect the sensitivity of intraoperative US to detect the presence and the extent of intrinsic brain tumours. Our findings could have translational potential as an intraoperative guidance tool.