The State of Machine Learning in Outcomes Prediction of Transsphenoidal Surgery: A Systematic Review.

The purpose of this analysis is to assess the use of machine learning (ML) algorithms in the prediction of postoperative outcomes, including complications, recurrence, and death in transsphenoidal surgery. Following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, we systematically reviewed all papers that used at least one ML algorithm to predict outcomes after transsphenoidal surgery. We searched Scopus, PubMed, and Web of Science databases for studies published prior to May 12, 2021. We identified 13 studies enrolling 5,048 patients. We extracted the general characteristics of each study; the sensitivity, specificity, area under the curve (AUC) of the ML models developed as well as the features identified as important by the ML models. We identified 12 studies with 5,048 patients that included ML algorithms for adenomas, three with 1807 patients specifically for acromegaly, and five with 2105 patients specifically for Cushing's disease. Nearly all were single-institution studies. The studies used a heterogeneous mix of ML algorithms and features to build predictive models. All papers reported an AUC greater than 0.7, which indicates clinical utility. ML algorithms have the potential to predict postoperative outcomes of transsphenoidal surgery and can improve patient care. Ensemble algorithms and neural networks were often top performers when compared with other ML algorithms. Biochemical and preoperative features were most likely to be selected as important by ML models. Inexplicability remains a challenge, but algorithms such as local interpretable model-agnostic explanation or Shapley value can increase explainability of ML algorithms. Our analysis shows that ML algorithms have the potential to greatly assist surgeons in clinical decision making.

Trends in Matriculation from Neurological Surgery Training Programs into Academic Versus Private Practice.

OBJECTIVE: A career in academic neurosurgery is an arduous endeavor. Specific factors influencing physician practice preferences remain unclear. This study analyzes data from the American Association of Neurological Surgeons membership identifying the impact of several demographic and educational characteristics influencing neurosurgical career choices centered on academia, private practice, or a combination in the United States. METHODS: A list of all current neurosurgeons was obtained from the American Association of Neurological Surgeons membership, and information on physician characteristics was collected via internet searches and institutional databases. The practice type of all neurosurgeons considered in this study were categorized as follows: private practice, academic, or a combination of private practice and academic, termed privademic. These data were subsequently correlated to race, gender, current age, training at a top 40 National Institutes of Health-funded medical school or residency program, and current practice. RESULTS: The median age of private practice and academic neurosurgeons was 58.18 and 53.61 years, respectively (P < 0.001). Age was significantly associated with practicing in an academic setting (odds ratio 0.96), with younger neurosurgeons pursuing careers in academia. Data indicated a positive and statistically significant contribution of female gender (P < 0.001) and training at a top-40 National Institutes of Health-funded institution to practicing in an academic setting (P < 0.01). CONCLUSIONS: Neurosurgery as a field has grown significantly over the past century. The authors recommend that future efforts seek to diversify the neurosurgical workforce by considering practice setting, demographic characteristics, and educational background.

Management of brain metastasis. Surgical resection versus stereotactic radiotherapy: a meta-analysis.

Background: Treatment of metastatic brain tumors often involves radiotherapy with or without surgical resection as the first step. However, the indications for when to use surgery are not clearly defined for certain tumor sizes and multiplicity. This study seeks to determine whether resection of brain metastases versus exclusive radiotherapy provided improved survival and local control in cases where metastases are limited in number and diameter. Methods: According to PRISMA guidelines, this meta-analysis compares outcomes from treatment of a median number of brain metastases ≤ 4 with a median diameter ≤ 4 cm with exclusive radiotherapy versus surgery followed by radiotherapy. Four randomized control trials and 11 observational studies (1693 patients) met inclusion criteria. For analysis, studies were grouped based on whether radiation involved stereotactic radiosurgery (SRS) or whole-brain radiotherapy (WBRT). Results: In both analyses, there was no difference in survival between surgery ± SRS versus SRS alone two years after treatment (OR 1.89 (95% CI: 0.47-7.55, P = .23) or surgery + WBRT versus radiotherapy alone (either WBRT and/or SRS) (OR 1.18 (95% CI: 0.76-1.84, P = .46). However, surgical patients demonstrated greater risk for local tumor recurrence compared to SRS alone (OR 2.20 (95% CI: 1.49-3.25, P < .0001)) and compared to WBRT/SRS (OR 2.93; 95% CI: 1.68-5.13, P = .0002). Conclusion: The higher incidence of local tumor recurrence for surgical patients suggests that more prospective studies are needed to clarify outcomes for treatment of 1-4 metastasis less than 4 cm diameter.

Robotic external ventricular drain placement for acute neurosurgical care in low-resource settings: feasibility considerations and a prototype design.

OBJECTIVE: Emergency neurosurgical care in lower-middle-income countries faces pronounced shortages in neurosurgical personnel and infrastructure. In instances of traumatic brain injury (TBI), hydrocephalus, and subarachnoid hemorrhage, the timely placement of external ventricular drains (EVDs) strongly dictates prognosis and can provide necessary stabilization before transfer to a higher-level center of care that has access to neurosurgery. Accordingly, the authors have developed an inexpensive and portable robotic navigation tool to allow surgeons who do not have explicit neurosurgical training to place EVDs. In this article, the authors aimed to highlight income disparities in neurosurgical care, evaluate access to CT imaging around the world, and introduce a novel, inexpensive robotic navigation tool for EVD placement. METHODS: By combining the worldwide distribution of neurosurgeons, CT scanners, and gross domestic product with the incidence of TBI, meningitis, and hydrocephalus, the authors identified regions and countries where development of an inexpensive, passive robotic navigation system would be most beneficial and feasible. A prototype of the robotic navigation system was constructed using encoders, 3D-printed components, machined parts, and a printed circuit board. RESULTS: Global analysis showed Montenegro, Antigua and Barbuda, and Seychelles to be primary candidates for implementation and feasibility testing of the novel robotic navigation system. To validate the feasibility of the system for further development, its performance was analyzed through an accuracy study resulting in accuracy and repeatability within 1.53 ± 2.50 mm (mean ± 2 × SD, 95% CI). CONCLUSIONS: By considering regions of the world that have a shortage of neurosurgeons and a high incidence of EVD placement, the authors were able to provide an analysis of where to prioritize the development of a robotic navigation system. Subsequently, a proof-of-principle prototype has been provided, with sufficient accuracy to target the ventricles for EVD placement.