Research On The Application Value Of Intelligent Multimodal Surgery In The Surgical Treatment Of High-grade Gliomas

Objective:At present,the primary treatment for high-grade gliomas（HGG）is to remove the tumor as far as possible in a safe range.However,HGG show infiltrating growth and have profound pathological"real"boundaries.At the same time,they are adjacent to many important nuclei,functional areas,entricles,White matter fiber conduction bundles and important intracranial blood vessels have brought difficulties to the surgical expansion of tumor removal.In the study,it was found that most gliomas have different high and low metabolism areas,and their high metabolism areas are highly invasive.In the past,we found that in the clinical practice of routine microscope-assisted surgical resection of glioma patients,some patients had reexamination imaging after surgery.There are leftover areas of high metabolism,and these leftovers may bring more poor prognosis to patients.With the development of multi-technology-assisted surgery,intelligent multi-modal surgery technology（preoperative multi-modal image fusion,intraoperative neuronavigation,intraoperative ultrasound,intraoperative neuroelectrophysiological monitoring,etc.）is adopted to assist the operation,and It is considered possible to remove the high metabolic area of the tumor,and then expand the tumor removal within the maximum safety range.This study explores the application value of intelligent multi-modal surgery in HGG surgical treatment.Methods:The clinical data of 88 patients whose pathology was confirmed to be HGG after neurosurgery in our hospital from January 2017 to August 2020 was retrospectively analyzed.Intelligent multi-modal technology surgery group（40 cases）using multi-modal imaging[Computed Tomography（CT）,Positron Emission Tomography Computed Tomography（PET-CT）,Magnetic resonance Imaging（MRI）,Perfusion Weighted Imaging（PWI）,Magnetic Resonance Specroscopy（MRS）,Diffusion Tensor Imaging（DTI）,etc.]Fusion technology to identify tumors high and low metabolic regions of interest（ROI）,intraoperative robotic navigation technology to mark ROI and obtain pathological specimens,navigation and intraoperative ultrasound guidance technology to determine the intraoperative high and low metabolic targets and tumor borders as much as possible,and ensure the removal of high-metabolic tumor areas under the microscope,and then expand the tumor removal within the maximum safety range.Postoperative pathology was used to observe and analyze the difference of Ki-67 expression in different regions of tumors of interest in high and low metabolism by immunohistochemical methods.Collect clinical data such as extent of resection（EOR）,craniotomy,postoperative complications,and functional status at 3months postoperatively.In addition,a group of 48 patients who underwent perioperative microscopy based on conventional MR was selected as the control group.The above-mentioned clinical prognostic data collection was also performed,and statistical data processing and analysis were performed.Results:（1）The operation progressed smoothly in all patients.Proficient in the application of multi-mode imaging,navigation and intraoperative ultrasound technology.According to multi-modal image fusion,mark ROI（tumor high metabolism,low metabolism,and peritumoral boundary）,and formulate surgical resection strategy,That is,ensure the removal of high metabolic areas and expand the tumor removal within a safe range.The robot navigation accurately targets and marks the high-metabolic area and the low-metabolic boundary around the tumor with methylene blue staining.The median registration error is1.13mm and the range is 0.44-2.00mm.Intraoperative ultrasound can realize part of real-time imaging navigation,which is of great value for real-time correction of navigation deviation and guidance of step-by-step tumor resection.There were no intraoperative complications in both groups.（2）The two groups were in gender,age,hemisphere involved,preoperative dyskinesia,preoperative epileptic seizures,tumor maximum diameter（cm）,tumor location,World Health Organization（WHO）classification,Karnofsky Performance Status（KPS）has no significant difference.The Gross Total Resection（GTR）of the cases in the intelligent multimodal surgery group was 85.0%（34/40）,while the GTR of the cases in the conventional microscope surgery group was 56.2%（27/48）,both GTR has a significant difference（P=0.004）.The area of the craniotomy bone flap in the intelligent multimodal surgery group was 25.4±9.4 cm²,and that of the conventional group was 33.0±10.2 cm².The area of the craniotomy bone flap in the former was smaller than that in the latter,and the difference was significant（P=0.001）.There was no statistically significant difference between the intelligent multimodal surgery group and the routine group in terms of postoperative complications of new-onset dyskinesia or aggravation,new-onset epilepsy or aggravation,hydrocephalus,speech disorder,subcutaneous fluid in the operation area,and intracranial infection.The median KPS score in the intelligent multimodal surgery group was90 at 3 months after surgery,ranging from 70 to 100;the median KPS score in the routine group was 90 at 3 months after surgery,ranging from 40 to 100.The postoperative functional prognosis of the intelligent multimodal surgery group was better than that of the conventional group,and the difference was significant（P=0.001）.In the intelligent multimodal surgery group,39 cases（97.5%）had no change or improvement in the KPS score after 3 months after the operation,and 1 case（2.5%）worsened;40 cases（83.3%）had no change or improvement in the routine group,8 cases（16.7%）were aggravated,and the difference between the two groups was statistically different（P=0.036）.（3）In this study,26targets in the high-metabolism zone were collected for pathology,and 30 targets in the low-metabolism zone around the tumor were selected for pathology,and the expression level of Ki-67 was detected by immunohistochemistry.Spearman correlation analysis between the pathological Ki-67 expression level of high and low metabolic targets and the corresponding PET-CT fluoro[F]deoxyglucose Standard Uptake Value(SUV_FDG)can be obtained:as the SUV_FDG increases,the corresponding Ki-67 expression level also increased,and the expression level of Ki-67 was positively correlated with SUV_FDG（r=0.803,P<0.05）.The expression level of Ki-67 in tumors with high metabolic ROI（49.8%±16.2）was significantly higher than the expression level of Ki-67 in tumors with low metabolic ROI（5.7%±3.4）,P<0.05.Conclusion:（1）Multi-modal image fusion based on metabolic and anatomical imaging can reflect the sensitivity of gliomas metabolic differences,which is highly consistent with the postoperative pathological proliferative differences（Ki-67 expression）.Intelligent fusion of preoperative images can be used as tumors.Evaluation of heterogeneity and biological behavior is conducive to formulating a reasonable surgical resection plan;（2）The application of auxiliary surgical techniques such as intraoperative precise navigation and intraoperative ultrasound guidance can accurately locate the key tumor resection target and the periphery of the tumor.The maximum precise removal of the active tumor（high proliferation and high metabolism）within the safety range significantly improves the effective tumor resection rate,effectively avoids the patient’s severe functional damage after surgery,improves the prognosis and delays tumor recurrence;（3）Based on the study of imaging hypermetabolism and pathological Heterogeneity,this study explored the significance of high-metabolism region recognition for the development of HGG surgical strategy with the aid of intraoperative precise localization techniques,in order to evaluate tumor proliferation Heterogeneity,understanding the biological characteristics of tumors,targeted surgical treatment and postoperative radiotherapy and chemotherapy and a series of clinical treatment plans to provide more valuable information.