Étienne Léger | Searcharxiv

arXiv2026

LARK: A Low-Cost, Accurate, Occlusion-Resilient, Kalman Filter-Assisted Tracking System for Image-Guided Surgery

Image-guided surgery (IGS) depends on accurate tracking of surgical instruments to provide real-time navigation relative to anatomical structures. Commercial stereo infrared trackers are accurate but prone to occlusion and cost-prohibitive for many settings. This work presents LARK, a multi-camera optical tracking system using commodity RGB hardware and multi-view redundancy and fusion. We develop and evaluate two complete tracking methods: multi-view monocular pose fusion and multi-view triangulation. Both methods are assessed under varying occlusion levels using a precision-machined grid and an anatomical head phantom, and compared against a gold-standard stereo infrared system. With five cameras and adaptive Kalman filtering, LARK achieves median target registration errors of 0.64 mm for point localization with triangulation and 0.73 mm for trajectory tracking with pose fusion on the machined grid. Camera-subset experiments show graceful degradation in adaptive pose-fusion accuracy as fewer views remain available. With tracking hardware costing under $1,000 USD, LARK provides a low-cost platform for image-guided surgery research. Hardware designs and software are publicly available at https://nist.mni.mcgill.ca/software/ , and datasets at https://nist.mni.mcgill.ca/data/ .

cs.CV↗

arXiv2025

A High-Fidelity Neurosurgical Training Platform for Bimanual Procedures: A Feasibility Study

Background. Bimanual psychomotor proficiency is fundamental to neurosurgical procedures, yet it remains difficult for trainees to acquire and for educators to objectively evaluate performance. In this study, we investigate the feasibility of a neurosurgical simulation platform that integrates an anatomically realistic brain model with surgical instrument tracking to support training and objective assessment of bimanual tasks in the context of subpial corticectomy. Methods. We developed and evaluated a neurosurgical simulation platform based on an ex-vivo calf brain model and a multi-camera tracking system capable of simultaneously capturing the motion of surgical instruments in both hands, including collection of real-time instrument trajectories and synchronized video recordings. These enabled extraction of motion-based, time-based, and bimanual coordination metrics. We conducted a case series involving 47 participants across four training levels: medical students, junior residents, senior residents, and neurosurgeons. Results. The tracking system successfully captured instrument motion during 81% of the periods when instruments were actively used throughout the simulation procedure. Several extracted metrics were able to significantly differentiate between levels of surgical expertise. In particular, instrument usage duration and custom-defined bimanual coordination metrics such as instrument tip separation distance and simultaneous usage time, show potential as features to identify participant expertise levels with different instruments. Conclusions. We demonstrated the feasibility of tracking surgical instruments during complex bimanual tasks in an ex-vivo brain simulation platform. The metrics developed provide a foundation for objective performance assessment and highlight the potential of motion analysis to improve neurosurgical training and evaluation.

cs.CE↗