Comparison of augmented reality and cutting guide technology in assisted harvesting of iliac crest grafts - A cadaver study.

BACKGROUND: Harvesting vascularized bone grafts with computer-assisted surgery represents the gold standard for mandibular reconstruction. However, current augmented reality (AR) approaches are limited to invasive marker fixation. This trial compared a markerless AR-guided real-time navigation with virtually planned and 3D printed cutting guides for harvesting iliac crest grafts. MATERIAL AND METHODS: Two commonly used iliac crest transplant configurations were virtually planned on 10 cadaver hips. Transplant harvest was performed with AR guidance and cutting guide technology. The harvested transplants were digitalized using cone beam CT. Deviations of angulation, distance and volume between the executed and planned osteotomies were measured. RESULTS: Both AR and cutting guides accurately rendered the virtually planned transplant volume. However, the cumulative osteotomy plane angulation differed significantly (p = 0.018) between AR (14.99 ± 11.69°) and the cutting guides (8.49 ± 5.42°). The cumulative osteotomy plane distance showed that AR-guided navigation had lower accuracy (2.65 ± 3.32 mm) than the cutting guides (1.47 ± 1.36 mm), although without significant difference. CONCLUSION: This study demonstrated the clinical usability of markerless AR-guided navigation for harvesting iliac crest grafts. Further improvement of accuracy rates might bring clinical implementation closer to reality.

Navigation of iliac crest graft harvest using markerless augmented reality and cutting guide technology: A pilot study.

BACKGROUND: Defects of the facial skeleton often require complex reconstruction with vascularized grafts. This trial elucidated the usability, visual perception and accuracy of a markerless augmented reality (AR)-guided navigation for harvesting iliac crest transplants. METHODS: Random CT scans were used to virtually plan two common transplant configurations on 10 iliac crest models, each printed four times. The transplants were harvested using projected AR and cutting guides. The duration and accuracies of the angulation, distance and volume between the planned and executed osteotomies were measured. RESULTS: AR was characterized by the efficient use of time and accurate rendition of preoperatively planned geometries. However, vertical osteotomies and complex anatomical settings displayed significant inferiority of AR guidance compared to cutting guides. CONCLUSIONS: This study demonstrated the usability of a markerless AR setup for harvesting iliac crest transplants. The visual perception and accuracy of the AR-guided osteotomies constituted remaining weaknesses against cutting guide technology.

Efficient Reject Options for Particle Filter Object Tracking in Medical Applications.

Reliable object tracking that is based on video data constitutes an important challenge in diverse areas, including, among others, assisted surgery. Particle filtering offers a state-of-the-art technology for this challenge. Becaise a particle filter is based on a probabilistic model, it provides explicit likelihood values; in theory, the question of whether an object is reliably tracked can be addressed based on these values, provided that the estimates are correct. In this contribution, we investigate the question of whether these likelihood values are suitable for deciding whether the tracked object has been lost. An immediate strategy uses a simple threshold value to reject settings with a likelihood that is too small. We show in an application from the medical domain-object tracking in assisted surgery in the domain of Robotic Osteotomies-that this simple threshold strategy does not provide a reliable reject option for object tracking, in particular if different settings are considered. However, it is possible to develop reliable and flexible machine learning models that predict a reject based on diverse quantities that are computed by the particle filter. Modeling the task in the form of a regression enables a flexible handling of different demands on the tracking accuracy; modeling the challenge as an ensemble of classification tasks yet surpasses the results, while offering the same flexibility.