Abstract
Abstract Gliomas are invasive tumors in which the extent of tumor resection is believed to correlate with prognosis. However, aggressive extraction risks deterioration due to functional tissue damage. One strategy for resolving this conflict is the technology of new modalities. The basis of this technology is digital transformation (DX), converting analog information called tacit knowledge on experiences into digital information for explicit knowledge. In the future, we will develop an AI system, generated by DX data, and support intraoperative decision-making with highly accurate future predictions. In order to realize this future-prediction surgery, we have developed a smart cyber operating theater (SCOT) with intraoperative MRI. More than 20 medical devices are connected by middleware (OPeLiNK®), and navigation spatial records and time-synchronized data are displayed on the strategy desk. A hyper version of SCOT has currently performed 277 surgeries, including 262 gliomas, 51 awake craniotomies, and 88 photodynamic therapies. For WHO grade 4 glioblastomas, we try to remove contrast-enhanced areas, and for grade 2/3 gliomas, to formulate an extraction strategy for each 3 genotypes, we perform preoperative AI prediction and intraoperative genetic diagnosis for genotype. Intraoperative flow cytometry is useful to predict the genotype (oligodendroglioma) and determine the extent of resection in some cases. Time-synchronized data showed points of surgical maneuvers causing speech arrest in awake surgery and decreases in MEP, and provided feedback during and after surgery. AI technology is also used for preoperative surgical simulation (GRID®), artifact reduction in iMRI, and brain shift prediction. The hyper SCOT supports the optimal removal of gliomas through intraoperative future prognosis prediction through device data and AI analysis. Furthermore, we are developing an AI robot (AIREC) that performs surgery through deep predictive learning, with the aim of commercialization after 2040. This work was partially supported by JST [Moonshot R&D] [Grant Number JPMJMS2031].
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have