Author ORCID Identifier

https://orcid.org/0000-0003-2214-3195

Date of Award

Winter 1-14-2026

Document Type

Thesis (Ph.D.)

Department or Program

Engineering Sciences

First Advisor

Keith D. Paulsen

Second Advisor

Xiaoyao Fan

Abstract

Modern surgical navigation systems improve procedural accuracy, limit complications, and can reduce reliance on intraoperative ionizing radiation, yet their adoption remains uneven across surgical domains. Informed by clinical immersion, this thesis identifies and addresses two unmet needs revealed through contrasting image-guidance practices in endovascular and spine surgery, where existing guidance techniques are costly, intermittent, or fundamentally limited. Specifically, these needs include (1) the absence of radiation-sparing navigation techniques for endovascular procedures, which rely heavily on 2D x-ray angiography, and (2) the need for dynamic registration methods in open spine surgery, where navigation accuracy can degrade after initial registration and intraoperative CT acquisition due to intraoperative spinal pose changes.

To address the first need, this thesis develops a fluoroscopy-free navigation approach for carotid endovascular surgery based on spatially tracked ultrasound (US) and image processing. As a prerequisite, robust and user-independent spatial US calibration methods are investigated, culminating in an image-based optimization technique that enables clinically accurate tracked US. This work establishes tracked ultrasound as a practical and adaptable quantitative 3D imaging modality for intraoperative navigation.

Building on this calibration framework, a radiation-sparing endovascular navigation approach is introduced and validated. Spatially tracked US is used to reconstruct vascular anatomy, register the surgical field to preoperative imaging, and localize endovascular instruments without reliance on fluoroscopy. Phantom experiments demonstrate high registration and instrument-localization accuracy, while a clinical data study quantifies preoperative-to-intraoperative arterial deformation and rigid registration performance of US reconstructions to preoperative CT angiography. Together, these results establish a navigation strategy that can extend beyond the carotid to broader peripheral artery applications.

To address the second need, this thesis advances dynamic registration techniques for open spine surgery. Building on the concept of stereovision-based surface reconstruction, this work explores a stereovision-to-stereovision registration paradigm that enables more frequent updates to spinal alignment without repeated intraoperative CT scans. Deep learning techniques for automatic bone-surface extraction are evaluated to provide consistent inputs required for reliable registration.

Collectively, these contributions advance radiation-sparing surgical navigation by enabling fluoroscopy-free endovascular guidance, improving robustness in tracked ultrasound calibration, and establishing a pathway toward more dynamic, non-radiative registration in spine surgery.

Original Citation

Warner, W.R., Bhattacharya, I., Evans, L.T., Mirza, S.K., Paulsen, K.D., Fan, X. (2026). Sparse-XM: Spine Pose Adjustment with RGB-D Bone Segmentation via Cross-Modality Label Transfer. In: Gee, J.C., *et al.* Medical Image Computing and Computer Assisted Intervention – MICCAI 2025. MICCAI 2025. Lecture Notes in Computer Science, vol 15968. Springer, Cham. https://doi.org/10.1007/978-3-032-05114-1_51

William R. Warner, Kirthi S. Bellamkonda, Richard J. Powell, Roberta M. diFlorio-Alexander, Keith D. Paulsen, Xiaoyao Fan, A novel three-dimensional navigation technology for ultrasound-guided transcarotid artery revascularization, JVS-Vascular Insights, Volume 3, 2025, 100260, ISSN 2949-9127, https://doi.org/10.1016/j.jvsvi.2025.100260. (https://www.sciencedirect.com/science/article/pii/S2949912725000777)

William R. Warner, Xiaoyao Fan, Ryan B. Duke, Haley E. Stoner, Chengpei Li, Songbai Ji, Linton T. Evans, Sohail K. Mirza, Keith D. Paulsen, "Automatic spine exposure segmentation in stereovision imaging using SAM 2 for driving image updating pipeline for preoperative to intraoperative registration in open spine procedures," Proc. SPIE 13408, Medical Imaging 2025: Image-Guided Procedures, Robotic Interventions, and Modeling, 1340804 (7 April 2025);~[https://doi.org/10.1117/12.3047443](https://doi.org/10.1117/12.3047443)~

William R. Warner, Xiaoyao Fan, Ryan B. Duke, Kristen L. Chen, Chengpei Li, Haley Stoner, Kirthi S. Bellamkonda, Linton T. Evans, Richard J. Powell, Sohail K. Mirza, Keith D. Paulsen, "Smart line detection and histogram-based approach to robust freehand ultrasound calibration," Proc. SPIE 12928, Medical Imaging 2024: Image-Guided Procedures, Robotic Interventions, and Modeling, 1292825 (29 March 2024);~[https://doi.org/10.1117/12.3006728](https://doi.org/10.1117/12.3006728)~

William R. Warner, Xiaoyao Fan, Ryan B. Duke, Kristen L. Chen, Haley Stoner, Chen Li, Shaoju Wu, Songbai Ji, Sohail K. Mirza, Keith D. Paulsen, "Towards accounting for intraoperative spine motion: a simulation study of registration between stereovision surfaces," Proc. SPIE 12466, Medical Imaging 2023: Image-Guided Procedures, Robotic Interventions, and Modeling, 124661H (3 April 2023);~[https://doi.org/10.1117/12.2654396](https://doi.org/10.1117/12.2654396)~

William R. Warner, Xiaoyao Fan, Ryan B. Duke, Tahsin M. Khan, Songbai Ji, Steven P. Baltic, Sohail K. Mirza, Keith D. Paulsen, "Preoperative-to-interoperative shift in spine pose measured as change in lordosis Cobb angle and its effect on navigational accuracy," Proc. SPIE 12034, Medical Imaging 2022: Image-Guided Procedures, Robotic Interventions, and Modeling, 120342B (4 April 2022);~[https://doi.org/10.1117/12.2612600](https://doi.org/10.1117/12.2612600)~

Download

Share

COinS