Objectives In managing a patient with glioblastoma multiforme (GBM), a surgeon must weigh up whether sufficient tumour can be removed so that the patient can enjoy the benefits of decompression and cytoreduction, without impacting on the patient’s neurological status. In a previous study we identified the five most important anatomical features on a pre-operative MRI that are predictive of surgical resectability and used them to develop a grading system. The aim of this study was to apply an artificial neural network (ANN) to improve the prediction of surgical resectability.

Methods A prospectively maintained database was searched between February and August 2017 to identify all adult patients with supratentorial GBM that underwent resection. Pre-operative MRI scans were scored using the aforementioned grading system and post-operative scans assessed to determine the extent of resection. Performance of the standard grading system and ANN were then evaluated by analysing their Receiver Operator Characteristic curves; Area Under Curve (AUC) and accuracy were calculated and compared using the t-­test with a value of p<0.05 considered significant.

Results In all, 47 patients were included, of which 18 (38.3%) were found to have complete excision. The AUC and accuracy were significantly greater using the ANN compared to the standard grading system (0.87 vs. 0.79 and 0.81 vs. 0.77 respectively; p<0.01 in both cases).

Conclusions An ANN allows for improved prediction of surgical resectability in patients with GBM.