Novel lateral fiducial marker positioning technique eliminates registration failures in high-body mass index patients during robot-assisted spine surgery Authors: Abhishek Soni, S. Vidyadhara, T. Balamurugan, Alia Vidyadhara Published: 2026 ABSTRACT ABSTRACTBackground:Unsuccessful intraoperative computed tomography (CT) registration creates significant technical barriers limiting robot-assisted spine surgery implementation in obese populations due to imaging system field-of-view constraints exceeding patient dimensions.Materials and Methods:This consecutive series evaluated 63 patients with body mass index exceeding 40 kg/m2 undergoing robot-assisted thoracolumbar fusion between October 2023 and November 2024. A lateral reference marker positioning strategy exploited O-arm’s asymmetric spatial coverage (40 cm transverse vs. 15 cm anteroposterior). Systematic preliminary radiographic assessment guided positioning selection, with lateral placement implemented when scout images demonstrated inadequate fiducial visualization or when patient dimensions approached field-of-view limitations.Results:Lateral positioning was implemented in 17 (27.0%) cases. This approach eliminated severe registration failures, with zero patients requiring more than one reattempt compared to 3 (6.5%) patients in the standard positioning cohort requiring two or more reattempts. Ultimate registration success reached 100% across all cases. First-attempt success was achieved in 82.5% of procedures. Patients requiring reattempts demonstrated significantly prolonged O-arm acquisition intervals (30.3 vs. 19.4 min, P = 0.003) and elevated radiation exposure (61.6 vs. 38.3 mGy, P < 0.001). Optimal pedicle screw positioning (Gertzbein–Robbins Grade A + B) was maintained regardless of technique (lateral 97.6% vs. standard 96.9%, P = 0.650).Conclusion:Lateral reference marker positioning resolves CT registration challenges in obese patients undergoing robot-assisted spine surgery, achieving universal success using existing equipment without additional capital investment. INTRODUCTION Robot-assisted spinal surgical systems have established superior pedicle screw insertion accuracy since initial regulatory approval in 2004.[123] Unsuccessful intraoperative computed tomography (CT) registration, however, creates substantial technical obstacles limiting platform implementation, especially among obese populations.[3456] When patient dimensions exceed the imaging system field-of-view during CT acquisition, fiducial markers cannot be visualized simultaneously with spinal anatomy, preventing registration and necessitating procedure conversion to fluoroscopic or freehand techniques. This incompatibility arises from asymmetric imaging coverage capabilities, with obesity constituting an independent failure predictor.[78] This consecutive cohort validates a lateral fiducial placement technique designed to overcome field-of-view limitations in patients exceeding 40 kg/m2 body mass index (BMI). MATERIALS AND METHODS Study design and setting This single-institution retrospective consecutive series spanned October 16, 2023, through January 31, 2025, adhering to the Declaration of Helsinki principles. The investigation compared outcomes between conventional midline fiducial placement and an adaptive lateral positioning strategy developed during the study interval. Institutional Ethics Committee approval was obtained, with informed consent waiver granted due to retrospective design and minimal risk profile. Strengthening the Reporting of Observational Studies in Epidemiology reporting guidelines were followed. All surgical procedures employed the MazorX Stealth Edition robotic platform (Medtronic, Minneapolis, MN) with intraoperative O-arm imaging utilizing Scan-and-Plan workflow. Patient selection Consecutive patients meeting the following criteria were included: robotic-assisted thoracolumbar fusion procedures, BMI exceeding 40 kg/m2, intraoperative O-arm scanning with Scan-and-Plan workflow, and postoperative CT imaging availability. Patients were excluded for: nonrobotic surgical approaches, alternative registration workflows such as CT-fluoroscopy alignment, BMI ≤40 kg/m2, abandoned robotic procedures, and absent postoperative CT imaging. Surgical protocol All interventions utilized the MazorX Stealth Edition robotic platform (Medtronic, Minneapolis, MN) with O-arm O2 imaging system (Medtronic, Minneapolis, MN) following the Scan-and-Plan workflow. Prone positioning was achieved on specialized bariatric spine tables (Allen Advance Table, Hill-Rom, Chicago, IL) with a weight capacity exceeding 272 kg. A standardized technical approach was established for high-BMI populations, addressing reference marker placement, robotic docking, preliminary scan verification, and imaging optimization. Comprehensive technical specifications are detailed in Supplementary Material 1. Fiducial marker positioning protocol Decision algorithm for lateral positioning Selection between midline and lateral flank marker placement followed methodical preliminary radiographic evaluation: Initial assessment: For all patients with BMI >40 kg/m2, perpendicular two-dimensional preliminary images (anteroposterior and lateral projections) were obtained with conventional midline fiducial placementScout image evaluation criteria: Lateral flank positioning was indicated when preliminary radiographs demonstrated any of the following: inability to visualize all four radio-opaque fiducial beads across both anteroposterior and lateral projections simultaneously, partial reference marker truncation at image boundaries, and combined patient anteroposterior thickness plus fiducial marker dimensions approaching the 15-cm sagittal field-of-view constraintBMI-based considerations: Although not representing an absolute threshold, lateral positioning was more frequently required among patients with BMI ≥45 kg/m2 (mean BMI: lateral 45.0 vs. standard 41.8 kg/m2, P < 0.001) and was universally implemented in superobese individuals (BMI ≥ 50 kg/m2, n = 2)Repositioning protocol: Upon detecting inadequate visualization on preliminary images, reference markers were repositioned laterally along the patient’s flank, confirmatory preliminary images were acquired, and three-dimensional acquisition proceeded utilizing a 40-cm field-of-view configuration. Initial assessment: For all patients with BMI >40 kg/m2, perpendicular two-dimensional preliminary images (anteroposterior and lateral projections) were obtained with conventional midline fiducial placement Initial assessment: For all patients with BMI >40 kg/m2, perpendicular two-dimensional preliminary images (anteroposterior and lateral projections) were obtained with conventional midline fiducial placement Scout image evaluation criteria: Lateral flank positioning was indicated when preliminary radiographs demonstrated any of the following: inability to visualize all four radio-opaque fiducial beads across both anteroposterior and lateral projections simultaneously, partial reference marker truncation at image boundaries, and combined patient anteroposterior thickness plus fiducial marker dimensions approaching the 15-cm sagittal field-of-view constraint Scout image evaluation criteria: Lateral flank positioning was indicated when preliminary radiographs demonstrated any of the following: inability to visualize all four radio-opaque fiducial beads across both anteroposterior and lateral projections simultaneously, partial reference marker truncation at image boundaries, and combined patient anteroposterior thickness plus fiducial marker dimensions approaching the 15-cm sagittal field-of-view constraint BMI-based considerations: Although not representing an absolute threshold, lateral positioning was more frequently required among patients with BMI ≥45 kg/m2 (mean BMI: lateral 45.0 vs. standard 41.8 kg/m2, P < 0.001) and was universally implemented in superobese individuals (BMI ≥ 50 kg/m2, n = 2) BMI-based considerations: Although not representing an absolute threshold, lateral positioning was more frequently required among patients with BMI ≥45 kg/m2 (mean BMI: lateral 45.0 vs. standard 41.8 kg/m2, P < 0.001) and was universally implemented in superobese individuals (BMI ≥ 50 kg/m2, n = 2) Repositioning protocol: Upon detecting inadequate visualization on preliminary images, reference markers were repositioned laterally along the patient’s flank, confirmatory preliminary images were acquired, and three-dimensional acquisition proceeded utilizing a 40-cm field-of-view configuration. Repositioning protocol: Upon detecting inadequate visualization on preliminary images, reference markers were repositioned laterally along the patient’s flank, confirmatory preliminary images were acquired, and three-dimensional acquisition proceeded utilizing a 40-cm field-of-view configuration. Standard positioning Reference markers were positioned centrally overlying the midline, placed as close to the skin surface as possible. Lateral positioning When preliminary radiographs indicated field-of-view constraints, reference markers were repositioned laterally along the patient’s flank, relocating the marker away from the central spine position toward the patient’s lateral aspect. This approach exploits O-arm’s asymmetric field-of-view capabilities (40-cm transverse coverage vs. 15-cm anteroposterior coverage) by positioning the marker laterally while maintaining simultaneous visualization of both spinal anatomy and fiducial beads within the scan volume. This positioning is illustrated in Figure 1. Lateral fiducial marker positioning for obese patients in robotic spine surgery: (a) marker placed laterally along the flank to sit closer to the spine; (b) lateral O-arm scout showing all four beads (red arrows) near the spine; (c) anteroposterior scout with incomplete marker/spine coverage; (d) three-dimensional computed tomography (40-cm field of view) confirming full coverage of the spine, marker, and both ilia Technical specifications Imaging parameters Standardized radiation parameters included 120.00 kilovoltage peak (kVp) and 186.25 milliampere-seconds (mAs) with low-to-medium dose selection, preventing fiducial marker oversaturation. Field-of-view selection employed a standard 20-cm diameter × 15-cm height for initial attempts, with a large 40-cm diameter × 15-cm height when lateral positioning was required. O-arm positioning was centered between the spine midline and the lateral marker location for optimal spatial coverage. Modifications for high-body mass index patients The protocol incorporated specific adaptations, including: (1) reference marker positioned maximally close to skin surface, (2) intraoperative determination for lateral marker repositioning based on preliminary image assessment, (3) conservative radiation parameters preventing marker oversaturation, and (4) systematic preliminary scan verification using perpendicular two-dimensional radiographs. Complete technical specifications and implementation protocols are detailed in Supplementary Material 1. Outcome measures Technical success and safety were evaluated using multiple metrics: Registration success metrics Ultimate registration success rate (successful registration achieved irrespective of attempt number), first-attempt registration success rate (successful automatic fiducial detection and registration on initial three-dimensional scan acquisition), and severe registration failures (≥2 reattempts or three or more total scan acquisitions). Procedural efficiency metrics Total operative duration (skin incision to closure), O-arm acquisition interval (from initial preliminary images to successful registration), and radiation exposure (measured in mGy). Safety and accuracy metrics Pedicle screw accuracy evaluated using Gertzbein–Robbins Grade A + B criteria on postoperative CT imaging[9] and perioperative complications. Registration failure classification Failure modes were categorized as: (1) inadequate field-of-view coverage, (2) fiducial marker oversaturation, (3) poor anatomical landmark visualization, or (4) technical/software issues. Statistical analysis Statistical analysis compared outcomes between positioning strategies (standard vs. lateral). Continuous variables are presented as mean ± standard deviation with ranges. Categorical variables are reported as frequencies and percentages with 95% confidence intervals (CIs). Comparisons between the groups (first-attempt success vs. registration reattempt, standard vs. lateral positioning, and BMI categories) were performed using Student’s t-test for continuous variables and Fisher’s exact test for categorical variables. Spearman’s rank correlation assessed relationships between BMI and surgical variables. Statistical significance was defined as P < 0.05. All analyses were performed using IBM SPSS Statistics for Windows, version 27 (IBM Corp., Armonk, N.Y., USA). Formal sample size calculations were not performed for this technical validation investigation. The study interval was designed to capture all eligible patients meeting the inclusion criteria during institutional implementation of the lateral positioning strategy. Given the exploratory nature of this technical modification and the absence of prior data on registration failure rates in this specific population (BMI >40 kg/m2), a pragmatic approach was adopted to include all consecutive cases over a defined interval to establish technical feasibility, safety, and efficacy across multiple outcome metrics. The 63 patients represent the complete institutional experience during the study period, providing a representative sample of high-BMI patients undergoing robotic spine surgery at this center. Fisher’s exact test was employed for categorical outcomes, including severe registration failures (defined as ≥2 reattempts). This analysis represents the complete consecutive cohort through January 2025. RESULTS Baseline characteristics and surgical details During the investigation period from October 2023 through January 2025, 63 consecutive patients with BMI exceeding 40 kg/m2 received robot-assisted spine surgery utilizing the MazorX Stealth Edition platform with intraoperative O-arm imaging. Baseline patient characteristics and surgical details appear in Table 1. Table 1Baseline characteristics and surgical detailsVariableOverall (n=63)Demographics Total patients, n63 Age (years), mean±SD61.5±12.7 Age range (years)15–79 Female, n (%)47 (74.6) Male, n (%)16 (25.4)Anthropometric measurements BMI (kg/m2), mean±SD43.6±2.70 BMI (kg/m2), range40.3–51.0 Class III obesity (BMI 40–49.9), n (%)61 (96.8) Super-obesity (BMI≥50), n (%)2 (3.2)Surgical details Emergency procedures, n (%)4 (6.3) Thoracic fusion, n (%)4 (6.3) Thoracolumbar fusion, n (%)4 (6.3) Lumbar fusion, n (%)51 (81.0) Lumbosacral fusion, n (%)4 (6.3) Instrumented levels, mean±SD2.0±1.85 Instrumented levels range1–9 Total pedicle screws inserted322SD - Standard deviation; BMI - Body mass index Baseline characteristics and surgical details SD - Standard deviation; BMI - Body mass index Registration Success Rates First-attempt O-arm registration succeeded in 52 of 63 cases (82.5%, 95% CI: 72.3%–90.0%). Ultimate registration success reached 100% across all 63 cases. Registration outcomes stratified by reference marker positioning strategy are detailed in Table 2. Table 2Registration success rates by reference marker positioning strategyOutcomeStandard positioning (n=46)Lateral positioning (n=17) P First-attempt success, n (%)36 (78.3)16 (94.1)0.26†Reattempts required, n (%)10 (21.7)1 (5.9)0.26†Mean reattempts per patient0.30±0.660.06±0.240.035‡Maximum reattempts31-Severe failures (≥2 reattempts), n (%)3 (6.5)0-Ultimate success rate, n (%)46 (100.0)17 (100.0)1.000†Fisher’s exact test; ‡Student’s t-test Registration success rates by reference marker positioning strategy †Fisher’s exact test; ‡Student’s t-test Within the standard positioning cohort, three patients experienced severe registration failures requiring two or more reattempts (6.5%), including one patient necessitating three reattempts (four total scan acquisitions). No patients in the lateral positioning cohort required more than one reattempt. Lateral fiducial positioning was implemented in 17 of 63 (27.0%) cases. Notably, lateral positioning eliminated severe registration failures: zero patients in this cohort required more than one reattempt, whereas 3 (6.5%) patients in the standard positioning cohort required two or more reattempts, including one patient requiring three reattempts (four total scan acquisitions). The mean reattempt frequency was lower with lateral positioning (0.06 vs. 0.30 reattempts per patient). First-attempt registration success showed no significant difference between BMI categories (P = 0.68). The single registration failure within the lateral positioning cohort occurred due to inadvertent use of the standard 20-cm field-of-view instead of the required 40-cm field-of-view protocol. This represents a protocol deviation rather than a technical limitation of the lateral positioning approach. Procedural efficiency and safety profile Procedural efficiency metrics and safety outcomes are presented in Table 3. Table 3Procedural efficiency and safety profileVariableFirst-attempt success (n=52)Registration reattempt (n=11) P Overall (n=63)Temporal metrics (min) Total operative duration185±76.7151±24.70.147179±71.5 O-arm acquisition interval19.4±8.9530.3±14.30.00321.0±10.7Safety metrics Radiation exposure (mGy)38.3±16.961.6±22.3<0.00141.9±19.6mGy - Milligray Procedural efficiency and safety profile mGy - Milligray Patients requiring registration reattempts demonstrated significantly prolonged O-arm acquisition intervals (30.3 min vs. 19.4 min; P = 0.003) and elevated radiation exposure (61.6 mGy vs. 38.3 mGy; P < 0.001) compared to those achieving first-attempt success. Total operative duration showed no significant difference between cohorts (185 min vs. 151 min; P = 0.147). Accuracy and complications Optimal pedicle screw positioning (Gertzbein–Robbins Grade A + B) was achieved in 315 of 322 screws (97.8%, 95% CI: 95.8%–99.0%), with no significant difference between positioning strategies (lateral: 97.6%, standard: 96.9%, P = 0.650). No major intraoperative complications occurred. All 11 cases requiring registration reattempts initially failed with standard central fiducial positioning and achieved successful registration upon reattempt, with 10 cases requiring repositioning to lateral fiducial placement. The single failure in the lateral positioning cohort was attributed to protocol deviation (use of 20-cm instead of 40-cm field-of-view). DISCUSSION Addressing a known technical barrier Registration failures represent a well-recognized obstacle in robotic spine surgery, often requiring conversion to alternative surgical techniques.[4] While robotic platforms generally demonstrate high accuracy, certain patient populations encounter persistent technical barriers to successful registration.[10] The growing obesity epidemic compounds these challenges, with obesity independently increasing surgical complexity, operative time, and complication rates.[111213] Our lateral fiducial positioning approach provides a validated solution to the spatial mismatch between obese patient anatomy and current imaging system capabilities. The elimination of severe registration failures with lateral positioning (0% vs. 6.5% requiring two or more reattempts) demonstrates the effectiveness of this technique in addressing a specific technical limitation. Technical mechanism The O-arm imaging system provides asymmetric spatial coverage, with 40 cm available in the lateral plane versus only 15 cm in the anteroposterior plane.[14] When a patient’s body thickness, combined with central marker placement, exceeds the 15-cm anteroposterior limit, registration becomes impossible. Lateral positioning relocates the reference marker to the patient’s flank, exploiting the expanded lateral coverage while maintaining simultaneous visualization of both spinal anatomy and fiducial markers. This geometric solution requires no additional equipment or capital investment, making it immediately applicable in centers with existing robotic platforms. Clinical outcomes This consecutive series achieved 100% ultimate registration success in 63 patients with BMI exceeding 40 kg/m2. Beyond technical feasibility, the approach improved procedural efficiency. Patients achieving first-attempt registration experienced significantly reduced O-arm acquisition intervals (19.4 vs. 30.3 min, P = 0.003), demonstrating that solving registration challenges streamlines the overall workflow. Surgical accuracy remained consistent regardless of positioning technique, with 97.8% optimal pedicle screw placement (Gertzbein–Robbins Grade A + B)[9] achieved across both lateral and standard approaches (lateral: 97.6% vs. standard: 96.9%, P = 0.650). These accuracy rates align with established robotic spine surgery outcomes,[15] confirming that lateral positioning maintains precision despite more challenging patient anatomy. Comparison with alternative approaches Several emerging technologies address registration challenges in complex cases. Ultrasound-based navigation systems offer radiation-free alternatives but require specialized equipment and prolonged validation.[16] Bed-mounted tubular retractors minimize soft-tissue deflection during navigation in obese patients.[17] Deep learning methods show potential but demand extensive computational infrastructure and large training datasets, limiting immediate clinical implementation.[18] Cone-beam CT systems demonstrate higher revision rates compared to O-arm systems (4.0% vs. 2.8%, P = 0.0081).[19] Lateral fiducial positioning offers practical advantages over these alternatives. The technique uses existing robotic platforms and imaging equipment without additional purchases, specialized tools, or extended validation periods. This makes it immediately implementable in any center performing robotic spine surgery with similar imaging systems. Safety and efficiency considerations Failed registration attempts requiring multiple scans constitute the primary source of excess radiation exposure in this population. Despite registration challenges in some cases, total operative time showed no significant difference between first-attempt success and reattempt groups (P = 0.147), indicating that the surgical workflow remains efficient even when imaging obstacles occur. Zero major complications occurred across all 63 procedures, demonstrating that the technique maintains safety standards while solving technical challenges. Implementation and study limitations The systematic protocol for lateral positioning includes several key elements: preliminary scan verification using perpendicular radiographs, proactive lateral positioning when field-of-view limitations are anticipated, and conservative radiation settings to prevent marker oversaturation. One registration failure in the lateral positioning group resulted from protocol deviation (using 20-cm instead of 40-cm field-of-view setting). This emphasizes that technique success requires strict protocol adherence, particularly regarding field-of-view selection. Team education during implementation is essential. Single-center experience limits generalizability, though these results demonstrate what can be achieved with consistent protocol implementation. The study period encompasses the institutional learning curve, providing realistic guidance for other centers. The investigation focused specifically on patients with BMI exceeding 40 kg/m2, limiting applicability to lower BMI ranges where standard positioning may suffice. While the technique was developed for the MazorX platform with O-arm imaging, the underlying geometric principles should apply to other robotic systems with similar field-of-view constraints. CONCLUSION Lateral reference marker positioning achieves 82.5% first attempt CT registration success among patients with BMI exceeding 40 kg/m2 undergoing robot-assisted spine surgery. This approach eliminates severe registration failures, with zero patients in the lateral cohort requiring more than one reattempt compared to up to three reattempts observed with conventional midline positioning. By resolving spatial incompatibility between enlarged patient dimensions and imaging system constraints, this technical modification enables previously excluded high-BMI individuals to access robotic surgical platforms. This investigation establishes immediately implementable protocols allowing spine surgeons to reliably perform robot-assisted procedures in obese populations. The technique minimizes conversion to fluoroscopic or freehand approaches, ensuring BMI alone does not restrict patient access to enhanced precision and safety benefits inherent to robotic spine surgery. Conflicts of interest There are no conflicts of interest.