OBJECTIVE Virtual reality (VR) and augmented reality (AR) systems are increasingly available to neurosurgeons. These systems may provide opportunities for technical rehearsal and assessments of surgeon performance. The assessment of neurosurgeon skill in VR and AR environments and the validity of VR and AR feedback has not been systematically reviewed. METHODS A systematic review following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines was conducted through MEDLINE and PubMed. Studies published in English between January 1990 and February 2021 describing the use of VR or AR to quantify surgical technical performance of neurosurgeons without the use of human raters were included. The types and categories of automated performance metrics (APMs) from each of these studies were recorded. RESULTS Thirty-three VR studies were included in the review; no AR studies met inclusion criteria. VR APMs were categorized as either distance to target, force, kinematics, time, blood loss, or volume of resection. Distance and time were the most well-studied APM domains, although all domains were effective at differentiating surgeon experience levels. Distance was successfully used to track improvements with practice. Examining volume of resection demonstrated that attending surgeons removed less simulated tumor but preserved more normal tissue than trainees. More recently, APMs have been used in machine learning algorithms to predict level of training with a high degree of accuracy. Key limitations to enhanced-reality systems include limited AR usage for automated surgical assessment and lack of external and longitudinal validation of VR systems. CONCLUSIONS VR has been used to assess surgeon performance across a wide spectrum of domains. The VR environment can be used to quantify surgeon performance, assess surgeon proficiency, and track training progression. AR systems have not yet been used to provide metrics for surgeon performance assessment despite potential for intraoperative integration. VR-based APMs may be especially useful for metrics that are difficult to assess intraoperatively, including blood loss and extent of resection.

Key Points Question What is the utility of a data set that contains videos of surgeons managing hemorrhage? Findings This quality improvement study of the Simulated Outcomes Following Carotid Artery Laceration (SOCAL), a public data set of surgeons managing catastrophic surgical hemorrhage in a cadaveric training exercise included 65 071 instrument annotations with recorded outcomes. Computer vision–based instrument detection achieved a mean average precision of 0.67 on SOCAL and a sensitivity of 0.77 and a positive predictive value of 0.96 at detecting surgical instruments from real intraoperative video. Meaning A corpus of videos of surgeons managing catastrophic hemorrhage is a novel, valuable resource for surgical data science.

No abstract

OBJECTIVE Experts can assess surgeon skill using surgical video, but a limited number of expert surgeons are available. Automated performance metrics (APMs) are a promising alternative but have not been created from operative videos in neurosurgery to date. The authors aimed to evaluate whether video-based APMs can predict task success and blood loss during endonasal endoscopic surgery in a validated cadaveric simulator of vascular injury of the internal carotid artery. METHODS Videos of cadaveric simulation trials by 73 neurosurgeons and otorhinolaryngologists were analyzed and manually annotated with bounding boxes to identify the surgical instruments in the frame. APMs in five domains were defined—instrument usage, time-to-phase, instrument disappearance, instrument movement, and instrument interactions—on the basis of expert analysis and task-specific surgical progressions. Bounding-box data of instrument position were then used to generate APMs for each trial. Multivariate linear regression was used to test for the associations between APMs and blood loss and task success (hemorrhage control in less than 5 minutes). The APMs of 93 successful trials were compared with the APMs of 49 unsuccessful trials. RESULTS In total, 29,151 frames of surgical video were annotated. Successful simulation trials had superior APMs in each domain, including proportionately more time spent with the key instruments in view (p < 0.001) and less time without hemorrhage control (p = 0.002). APMs in all domains improved in subsequent trials after the participants received personalized expert instruction. Attending surgeons had superior instrument usage, time-to-phase, and instrument disappearance metrics compared with resident surgeons (p < 0.01). APMs predicted surgeon performance better than surgeon training level or prior experience. A regression model that included APMs predicted blood loss with an R2 value of 0.87 (p < 0.001). CONCLUSIONS Video-based APMs were superior predictors of simulation trial success and blood loss than surgeon characteristics such as case volume and attending status. Surgeon educators can use APMs to assess competency, quantify performance, and provide actionable, structured feedback in order to improve patient outcomes. Validation of APMs provides a benchmark for further development of fully automated video assessment pipelines that utilize machine learning and computer vision.

Graph Neural Networks (GNNs) are a popular technique for modelling graphstructured data that compute node-level representations via aggregation of information from the local neighborhood of each node. However, this aggregation implies increased risk of revealing sensitive information, as a node can participate in the inference for multiple nodes. This implies that standard privacy preserving machine learning techniques, such as differentially private stochastic gradient descent (DP-SGD) -which are designed for situations where each data point participates in the inference for one point only -either do not apply, or lead to inaccurate solutions. In this work, we formally define the problem of learning 1-layer GNNs with node-level privacy, and provide an algorithmic solution with a strong differential privacy guarantee. Even though each node can be involved in the inference for multiple nodes, by employing a careful sensitivity analysis and a non-trivial extension of the privacy-by-amplification technique, our method is able to provide accurate solutions with solid privacy parameters. Empirical evaluation on standard benchmarks demonstrates that our method is indeed able to learn accurate privacy preserving GNNs, while still outperforming standard non-private methods that completely ignore graph information.