Health & Medicinepreprint2026-11-20

Markerless three-dimensional biomechanical analysis of manual flexible ureteroscopy: experience-related differences in wrist load during lower-calyx access

Open access0 citations

Abstract

Purpose To quantify upper-limb kinematics and estimated biomechanical load (a kinematics-derived load proxy, not measured muscle force) during simulated manual flexible ureteroscopy (fURS) using markerless 3D pose estimation, testing whether experience-related differences concentrate in lower-calyx access. Methods Eighteen urologists (10 experts, 8 novices) performed a phantom task: renal pelvis to upper, middle and lower calyx, returning to the pelvis between calyces. 3D landmarks from smartphone video (MediaPipe Pose) were converted to joint angles, jerk and inverse-dynamics load estimates. Primary outcomes were right-wrist force, jerk and angle in the lower-calyx phase; others were exploratory, analysed with participant-level Welch tests, Hedges g, 95% CIs, Holm correction and a mixed model for group × phase interaction. Results Procedure time was longer in novices (115.5 vs 61.5 s). In the lower-calyx phase, novices showed higher right-wrist force (1.87 vs 1.16 N; Holm p = 0.026), jerk (Holm p = 0.020) and angle (31.6° vs 24.9°; Holm p = 0.026), with no upper-calyx differences; force and jerk showed significant group × phase interactions (p = 0.004, 0.007) but angle did not (p = 0.35). Force, jerk and dispersion correlated strongly with movement speed (ρ = 0.93–0.99), and group differences were attenuated after speed adjustment. Conclusion In this simulator study, experience-related differences in estimated wrist load emerged mainly during lower-calyx access, whereas wrist angle differed throughout, pinpointing an ergonomic target. Because estimates were kinematics-derived and not EMG-validated, findings are hypothesis-generating; correlating them with physiological muscle activity is an important next step for robot-assisted ureteroscopy design.

// Source

View paper (DOI)Open access versionOpenAlexResearch SquarePublished 2026-11-20

Authors: Satoshi Kobayashi, Keiji Tsukino, Mikifumi Koura, Masaki Shiota

Institutions: Kyushu University