Artificial Intelligence in Brain Tumour Surgery—An Emerging Paradigm

Artificial intelligence (AI) platforms have the potential to cause a paradigm shift in brain tumour surgery. Brain tumour surgery augmented with AI can result in safer and more effective treatment. In this review article, we explore the current and future role of AI in patients undergoing brain tumour surgery, including aiding diagnosis, optimising the surgical plan, providing support during the operation, and better predicting the prognosis. Finally, we discuss barriers to the successful clinical implementation, the ethical concerns, and we provide our perspective on how the field could be advanced.

Prediction of Glasgow Outcome Scale after Barbiturate Coma Therapy in patients suffering from Refractory Intracranial Hypertension after Brain Tumour Surgery: Incorporating Ensemble approach into automated Machine Learning

Ensemble approach is incorporated into Automated machine learning to predict Glasgow outcome scale and In-hospital mortality in patients receiving Barbiturate coma therapy for refractory intracranial hypertension post Brain tumour surgery.

Surgical site infection in paediatric posterior fossa surgery: does pathology matter?

Objectives The aim of this study was to explore the rates and potential risks of surgical site infection (SSI) after posterior fossa surgery for tumour resection in children. Methods We retrospectively reviewed our local paediatric (age < 16 years) database for all cases of posterior fossa (PF) brain tumour surgery between November 2008 and November 2019. We collected patient demographics, tumour histology/location, and the event of postoperative surgical site infection. Results Overall, 22.1% (n=15) developed SSI out of sixty-eight children undergoing PF surgery for resection of brain tumours; 73.3% of them had a confirmed diagnosis of medulloblastoma. There was no statistically significant difference in the age (5.1 ± 0.60 vs. 6.2 ± 0.97 years; p=0.47) and duration of operation (262 vs. 253 min; p = 0.7655) between the medulloblastoma group and other tumours. Although the rate of postoperative hydrocephalus was higher in the medulloblastoma group (12.9% vs. 0%), this was not associated with increased SSI. Rates of CSF leak between the 2 groups were not different. Conclusion Medulloblastoma as a pathological entity seems to carry higher risk of postoperative surgical site infection compared to other types of paediatric posterior fossa tumours. Further larger studies are required to look into this causal relationship and other risk factors that might be involved.

Long-term motor deficit in brain tumour surgery with preserved intra-operative motor-evoked potentials

Muscle motor-evoked potentials are commonly monitored during brain tumour surgery in motor areas, as these are assumed to reflect the integrity of descending motor pathways, including the corticospinal tract. However, while the loss of muscle motor-evoked potentials at the end of surgery is associated with long-term motor deficits (muscle motor-evoked potential-related deficits), there is increasing evidence that motor deficit can occur despite no change in muscle motor-evoked potentials (muscle motor-evoked potential-unrelated deficits), particularly after surgery of non-primary regions involved in motor control. In this study, we aimed to investigate the incidence of muscle motor-evoked potential-unrelated deficits and to identify the associated brain regions. We retrospectively reviewed 125 consecutive patients who underwent surgery for peri-Rolandic lesions using intra-operative neurophysiological monitoring. Intraoperative changes in muscle motor-evoked potentials were correlated with motor outcome, assessed by the Medical Research Council scale. We performed voxel–lesion–symptom mapping to identify which resected regions were associated with short- and long-term muscle motor-evoked potential-associated motor deficits. Muscle motor-evoked potentials reductions significantly predicted long-term motor deficits. However, in more than half of the patients who experienced long-term deficits (12/22 patients), no muscle motor-evoked potential reduction was reported during surgery. Lesion analysis showed that muscle motor-evoked potential-related long-term motor deficits were associated with direct or ischaemic damage to the corticospinal tract, whereas muscle motor-evoked potential-unrelated deficits occurred when supplementary motor areas were resected in conjunction with dorsal premotor regions and the anterior cingulate. Our results indicate that long-term motor deficits unrelated to the corticospinal tract can occur more often than currently reported. As these deficits cannot be predicted by muscle motor-evoked potentials, a combination of awake and/or novel asleep techniques other than muscle motor-evoked potentials monitoring should be implemented.

SURG-12. PAEDIATRIC BRAIN TUMOUR SURGERY: HOW CAN WE REPORT OUR SURGICAL OUTCOMES AND OPERATIVE MORBIDITY?

OBJECTIVE Our objective was to quantify resection outcomes and operative morbidity in paediatric brain tumour surgery using existing scales, assessing their applicability. METHODS We investigated morbidity using the Clavien-Dindo (CD) scale and the Drake classification. All paediatric patients receiving a biopsy or craniotomy for an intracranial tumour in a single tertiary paediatric neurosurgery centre between January 2008 and December 2018 were studied. Complications up to day 30 post op were graded. RESULTS There were 459 operations: 92 biopsies and 367 craniotomies comprising 166 infratentorial and 292 supratentorial tumours. Median age was 9 years (56% male). The surgical goal was achieved or exceeded in 94% of cases. Thirty-day mortality was 1.31% with all deaths related to disease and none to surgical complications. The overall CD score was 1 in 10.9% of cases, 2 in 18.9%, 3A in 1.7%, 3B in 11.8%, and 4 in 1.1%. There was no operative morbidity in 54% of cases. Using the Drake classification, meningitis was seen in 3.92% of cases, seizures in 3.92%, neurological deficit (that persisted at 30 days) in 8.5%, CSF leak in 5.01%, wound infection in 1.96%, haemorrhage 1.75 %, shunt infection in 1.53%, shunt block in 0.65%, medical complications in 2.4%, and others in 3.05%. CONCLUSIONS This is the largest series presenting morbidity from paediatric brain tumour surgery, and the first to validate the CD scale. Our morbidity on the Drake scale was comparable with other series. There is a need to develop improved tools to quantify morbidity in this high-risk specialty.

A COMPARISON OF 3% HYPERTONIC SALINE AND MANNITOL FOR BRAIN RELAXATION DURING ELECTIVE SUPRATENTORIAL BRAIN TUMOUR SURGERY

Brain tumours may cause disturbances of cerebral auto-regulation & the blood brain barrier (BBB). These effects could possibly have an unfavourable influence on brain condition. To compare the effects of the application of a bolus hypertonic saline with those of mannitol on intraoperative brain relaxation. Ninety patients of ASA grade II and III in the age group >18 years with supratentorial brain tumours posted for elective brain tumour surgery were studied in the current study. They were randomly allocated into two groups. Each group enrolled with 45 patients- Group HTS : receiving 160 ml of 3% hypertonic saline. Group M : receiving 150 ml of 20% mannitol. Patients from both groups were induced with injection fentanyl, propofol & rocuronium. Maintained with N2O: O2 in a ratio of 2:1 combined with isoflurane, fentanyl and vecuronium. End tidal Co2 pressure were maintained between 35-40 mm Hg. Arterial blood pressure were kept within baseline value ± 20. Hence application of hypertonic saline is a more effective method for brain relaxation in patients posted for elective supratentorial brain tumours surgery and can be a great value in routine practice.