European Spine Journal - 2026-07-13 - Journal Article
Optimizing the mathematical model and technical standards for unilateral biportal endoscopic spinal surgery through machine learning-based video analysis.
Cheng D, Wu H, Zhou C, Wang W, Li N, Shao C, Wang H
Topics
Key Takeaway
A machine learning-derived predictive formula for UBE portal placement reduced operative time and intraoperative radiation exposure, with greatest benefit in patients with BMI >28 or BMI <18.5, though the specific magnitude of time reduction is not reported in the abstract.
Summary Depth
Choose how much analysis to show on this article page.
Summary
This study developed and validated a predictive formula for UBE portal incision placement and interportal distance using machine learning-assisted video analysis of 105 patients stratified into pilot (n=35), control (n=35), and experimental (n=35) groups. The experimental group using the algorithmic formula demonstrated reduced operative time, decreased intraoperative fluoroscopy exposure, and improved surgeon-rated ergonomics compared to controls. Benefits were most pronounced at BMI extremes (>28 or <18.5), and the optimal endoscopic working angle was defined as 30°–50°.
Key Limitation
The predictive formula itself is not presented in readable form in the abstract, making independent validation or clinical adoption impossible without access to the proprietary algorithm.
Original Abstract
OBJECTIVE
To enhance the efficacy of the unilateral biportal endoscopic spinal procedure, the spinal triangle concept is proposed, integrating a digital model to establish standardized reference parameters and technical protocols for biportal positioning and incision delineation.
METHODS
A retrospective-prospective cohort of 105 patients undergoing UBE for lumbar discectomy and unilateral spinal canal decompression between March 2024 and February 2026 was analyzed. The cohort was stratified into three groups: 35 cases for pilot testing, 35 as controls, and 35 as the experimental group where preoperative incision parameters were calculated using a proprietary algorithmic formula. Surgical procedures were executed and meticulously documented, with subsequent manual data annotation, machine learning-assisted computational analysis, and mathematical modeling to enhance the UBE spinal triangle framework. Surgical outcomes were assessed by attending surgeons via a standardized Likert scale, while patient-reported outcomes (VAS and ODI scores) were systematically recorded at postoperative intervals. Quantitative video-derived metrics underwent rigorous statistical evaluation, including Student's t-tests, multivariate linear regression, logistic regression analysis, and Pearson's chi-square tests.
RESULTS
The optimal surgical angle was determined to be within the range of 30°-50°. A standardized predictive formula for incision positioning and interportal distance was established: [Formula: see text]. The experimental group utilizing this formula demonstrated a significant reduction in operative duration and enhanced surgical ergonomics under comparable BMI conditions, with particularly pronounced benefits observed in patients with BMI > 28 or BMI < 18.5. Personalized incision placement and optimized interportal distance contributed to decreased operative time, minimized intraoperative radiation exposure, and improved procedural fluency and surgeon satisfaction.
CONCLUSION
In UBE procedures, the existence of an endoscopic triangular working zone was confirmed. The implementation of a standardized predictive formula for incision positioning and interportal distance enhances surgical efficiency and methodological precision.