E-199 AI-driven analytics platform for advanced surgical or monitoring in neurointervention: World’s first feasibility study
Bibliographic record
Abstract
Introduction/Purpose Safety in the Operating Room (OR) has improved in recent years with technical advancements, from modernization of medical equipment to personnel training and implementation of standard operating procedures. While the operator and surgical team have been trained with communication and teamwork skills that support situational awareness of errors in the OR, adverse events and workflows have nevertheless gone unnoticed. The OR Black Box™ is a system that tracks activities in the OR with cameras, sensors, and audio recording equipment from which data can be analyzed by deep learning technology. There is opportunity to proactively evaluate safety and efficiency in the interventional radiology workflow for endovascular interventions. The purpose of this study is to assess the angiosuite environment and identify potential areas of improvement in the neurointerventional radiology suite to support the clinical team in improving patient care. Materials and Methods The OR Black Box™ system was installed in two dedicated neurointervention procedure suites: Philips Azurion and Siemens ARTIS icono. Audiovisual data was securely transmitted from each OR to the Surgical Safety Technologies Servers at the hospital data center. Analysis of the technical and non-technical performance, as well as environmental distractions during the procedure was conducted using a machine learning algorithm. Technical performance was assessed with the Generic Error Rating Tool (GERT) and Objective Structured Assessment of Technical Skills (OSATS). The non-technical performance was assessed by the Scrub Practitioners’ intra-operative non-technical skills (SPINTS - SPLINTS) and Non-Technical Skills for Surgeons (NOTSS). Results We present data on the initial series of cases performed in our two angiosuites, including procedures for aneurysms, acute ischemic stroke, and arteriovenous embolization. The workflow performance will be analyzed to identify organizational and patient-related disruptions, as well communication gaps that limit team cohesion for optimal patient care. Furthermore, the results will demonstrate the gaps in the surgical team’s awareness of adverse events that could be addressed by improving the physical organization of the operating room and communication method between the neurointerventional, nursing, technologist, and anesthesiology teams. As such, the output of the OR Blackbox™ will reveal the root cause of adverse outcomes that can be prevented through individualized team training interventions for each of the two procedure suites. Results will be discussed in comparison to OR Blackbox™ in other areas of surgery, which would encourage adaptations of quality improvement practices between disciplines. Conclusion The current study will provide insight on the implementation of the OR Black Box™ in the first neurointerventional radiology suites in the world. The preliminary results will guide implementation of improved protocols to optimize OR safety and increase efficiency of the interventional neurosugery workflow. We believe that deep learning technology will drive the future of neurointerventional procedures, and neurointerventional safety is no exception. While new understandings could disrupt standard practices, the ultimate safety and efficiency achieved will optimize patient outcomes. Disclosures V. Chan: None. N. Cancelliere: None. V. Mendes Pereira: None.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.004 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.007 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".