Response to the letter to the editor: Augmented reality-guided pedicle screw fixation: an experimental study

Article information

Asian Spine J. 2025;19(5):875-876

Publication date (electronic) : 2025 October 28

doi : https://doi.org/10.31616/asj.2025.0533.r2

Sang-Min Park^,1, Dongjoon Kim², Jiwon Park³, Ho-Joong Kim¹, Jin S. Yeom¹

¹Spine Center, Department of Orthopaedic Surgery, Seoul National University Bundang Hospital, Seoul National University College of Medicine, Seongnam, Korea

²School of Information Convergence, College of Software and Convergence, Kwangwoon University, Seoul, Korea

³Department of Orthopaedics, Korea University Ansan Hospital, Korea University College of Medicine, Ansan, Korea

Corresponding author: Sang-Min Park Spine Center, Department of Orthopaedic Surgery, Seoul National University Bundang Hospital, Seoul National University College of Medicine, 82 Gumi-ro 173beon-gil, Bundang-gu, Seongnam 13620, Korea Tel: +82-31-787-7208, Fax: +82-31-787-4056, E-mail: psmini@naver.com

Received 2025 August 28; Accepted 2025 September 5.

To the Editor,

We thank the correspondent for their thoughtful and constructive comments on our study regarding augmented reality-guided pedicle screw fixation [1]. The statistical limitations and methodological concerns raised are valid and merit careful consideration. We appreciate the opportunity to respond to these important points.

We acknowledge that our sample size of 50 screws across five specimens is limited for distinguishing between inter-specimen and intra-specimen variability. This represents a recognized limitation of our pilot study design. Future investigations will incorporate larger sample sizes with multiple surgeons to enable proper statistical analysis using mixed-effects models or similar approaches to account for hierarchical data structure [2].

While we recognize the anatomical differences between porcine and human spines, the porcine lumbar model remains a well-established and validated model for pedicle screw research [3,4]. Comparative anatomical studies have demonstrated substantial similarities in pedicle dimensions and trabecular bone structure between porcine and human lumbar vertebrae [5]. Nevertheless, we fully acknowledge the need for progression to human cadaveric studies and ultimately clinical validation to establish translational relevance.

The absence of bleeding, patient movement, and complex pathological conditions represents a deliberate design choice for this proof-of-concept study. As an initial technology assessment, we sought to establish baseline accuracy under controlled conditions before introducing additional variables. However, we recognize that real-world factors will significantly impact system performance, and future studies must systematically evaluate these conditions, including dynamic accuracy assessment during simulated patient movement and performance in pathological bone conditions.

The concern regarding “outlier effects” is well-taken. While our mean deviations were submillimeter, reporting only aggregate data may indeed mask clinically significant outliers. Future publications will include complete accuracy distributions, maximum deviation values, and worst-case scenario analysis. We agree that clinical safety margins should be emphasized over average accuracy metrics, particularly given that even minor breaches can have serious consequences in critical anatomical regions [6].

The 2.2-minute placement time represents only the active screw insertion phase and does not reflect the complete workflow including system setup, registration, and planning phases. A comprehensive time-motion study incorporating all procedural steps is essential for meaningful efficiency evaluation. The point-pair matching technique using 15 landmarks may indeed introduce systematic registration errors that warrant further investigation to minimize user-dependent variability.

The correspondent raises several critical research questions that represent essential next steps for augmented reality (AR)-assisted pedicle screw fixation (ARPSF) development. We anticipate that osteoporotic bone and spinal deformities will present significant challenges, particularly for registration accuracy in patients with severe anatomical distortion. We hypothesize that experienced spine surgeons may demonstrate minimal error rates with ARPSF technology, while surgeons with limited experience may benefit from accelerated skill acquisition through intuitive AR guidance [7,8].

Comparative studies with existing navigation modalities will likely demonstrate distinct advantages of AR guidance [9]. Unlike conventional systems requiring attention shifts to external monitors, ARPSF offers more intuitive assistance while maintaining surgical focus within the operative field, potentially enabling more stable screw fixation. However, cost-effectiveness analysis remains uncertain and will require systematic evaluation through extended clinical use to determine economic impact and long-term outcomes.

Our study represents an initial step in evaluating AR technology for pedicle screw placement. While the limitations identified are substantial and must be addressed through systematic follow-up investigations, we believe the demonstrated submillimeter accuracy provides sufficient proof-of-concept to justify continued development. The correspondent’s comments will guide our future research design to ensure more robust evidence generation. We remain committed to rigorous scientific evaluation of this technology before clinical implementation and appreciate the constructive dialogue that advances the field.