A neurosurgical navigation system based on intraoperative tumour remnant estimation.

This paper proposes a method to intra-operatively visualize the process of tumour resection until complete resection is accomplished. A fuzzy connectedness method that is robust against image noises was used to identify the tumour position and volume. Based on the tumour segmentation results, the removed area and the residual tumour tissues were examined with reference to the electrocautery trace log. Unique processes that are specific to glioma resection were introduced in the method to improve the accuracy of estimation. Invalid data in the trace log were excluded, and the tumour region surrounded by valid log points was included in the removed area. The proposed system also produces an alarm to indicate whether the electrocautery is being accurately performed within the tumour area. Thus, this surgical navigation system can assist surgeons in intuitively monitoring the tumour resection process and properly removing tumour remnants. Preliminary experiments and a clinical pilot study showed the feasible application of this method.