Utilization of Magnetic Resonance Imaging–based Synthetic Computed Tomography for Pedicle Screw Navigation in Children: A Feasibility Study☆ Authors: Jaysson T. Brooks, Judy-Mae Lima, Emmanuel Ogu, David Thornberg, Daniel J. Sucato, Amy McIntosh, Megan Johnson, Brandon Ramo, Vivek Kalia, J. Herman Kan Published: 2026 ABSTRACT BackgroundPedicle screw instrumentation using a fluoroscopy-assisted freehand technique (FAFT) or using navigation based on a preoperative traditional computed tomography (tCT) scan is a common practice. However, concerns remain regarding the long-term effects of repeated ionizing radiation exposure. In 2025, the first pediatric patients in the USA underwent navigated posterior spinal fusions (PSFs) using magnetic resonance imaging (MRI) converted to CT scan–like quality, known as synthetic CT (sCT). The purpose of this study is to report on the clinical, radiographic, and complication outcomes of this cohort and compare radiation exposure between sCT, tCT, and FAFT modalities.MethodsTwenty patients received PSFs using sCT imaging and machine-vision image-guided surgical (MvIGS) navigation. Fluoroscopy was used in sCT cases to confirm the location of T12 and for small adjustments in alignment after rod insertion. To compare radiation exposure, the patients who received sCT were compared with a matched cohort of 135 patients receiving PSFs using the FAFT and 27 patients receiving PSFs using tCT imaging paired with MvIGS navigation. Radiation exposure was measured in mGy. Pedicle morphologies were measured and classified according to Sarwahi et al.ResultsTwelve patients (63%) who received sCT had an idiopathic diagnosis. Mean age at surgery was 13.6 ± 4 years. Across 335 pedicles instrumented, 13% were type C or D in morphology. The mean radiation exposure to the 27 patients receiving a preoperative tCT scan combined with routine intraoperative fluoroscopy was 5.44 mGy. For the 20 sCT cases, the mean radiation exposure was 0.90 mGy, reflecting an 83.5% reduction in radiation (P = .001). In comparison, 3 surgeons utilized FAFT during PSFs, producing a mean radiation dose of 7 mGy. Thus, navigated sCT cases resulted in an 87.1% reduction in radiation exposure in comparison to FAFT cases (P < .001). Within the sCT cohort, there were no intraoperative neuromonitoring (IONM) changes, cerebrospinal fluid (CSF) leaks, or postoperative neurologic deficits.ConclusionsNavigated PSFs utilizing sCTs paired with MvIGS navigation allow for instrumentation of small pedicles with an 83.5% reduction in radiation compared with tCTs and an 87.1% reduction in radiation compared with FAFT cases.Key Concepts(1)In this study, synthetic CTs are created by taking MRIs and converting this imaging into CT-scan-like quality using a proprietary, artificial intelligence mediated algorithm.(2)Synthetic CTs, paired with MvIGS navigation, were used to navigate pedicle screws in idiopathic, syndromic, congenital, and neuromuscular scoliosis diagnoses.(3)Difficult type C and D pedicles were instrumented using MvIGS navigation based on sCT imaging.(4)There is an 83.5% reduction in radiation exposure to the patient for MvIGS navigation cases using synthetic CTs compared with MvIGS navigation cases using traditional CTs.(5)In comparison to patients receiving PSFs using a fluoroscopy-assisted freehand technique, MvIGS navigation cases using synthetic CTs resulted in an 87% reduction in radiation exposure for both the surgeon and the patient.Level of EvidenceIII, Retrospective Cohort Study Introduction A commonly used method for pedicle screw instrumentation during a posterior spinal fusion (PSF) is the fluoroscopy-assisted freehand technique [1] (FAFT), which uses intraoperative fluoroscopy to identify and guide the pedicle screw trajectory. While widely available and familiar to most surgeons, FAFT often requires repeated fluoroscopic imaging, contributing to ionizing radiation exposure for patients, surgeons, and operative staff [2]. In patients with adolescent idiopathic scoliosis (AIS), cumulative radiation exposure has been associated with a markedly increased lifetime cancer risk, reported as up to a five-fold increase compared with the general population [3]. As a result, there is growing interest in techniques that maintain precision during pedicle screw instrumentation while minimizing radiation exposure across the patient's treatment course. Machine-vision image-guided surgery (MvIGS) is a spinal navigation technology that utilizes high-resolution stereoscopic cameras and structured visible light to construct a three-dimensional topographical map of the patient's exposed osseous surface anatomy, which is then coregistered to a preoperative traditional computed tomography (tCT) data set [4,5]. Unlike conventional navigation platforms that rely on intraoperative fluoroscopy or tCT for registration, MvIGS acquires over 250,000 surface data points at a rate exceeding 52,000 points per second, enabling rapid surface-to-image registration with a mean registration time of approximately 41 seconds compared with 258-794 seconds for fluoroscopy-based navigation systems. These MvIGS navigation systems improve the safety and accuracy of pedicle screw instrumentation while reducing fluoroscopy time relative to traditional freehand or FAFT methods [[6], [7], [8], [9]]. While accurate, the MvIGS navigation system still relies on preoperative tCT scans for registration, which remains a significant source of ionizing radiation to the patient [4]. Synthetic CT (sCT) has emerged as a potential solution to this limitation. sCT leverages artificial intelligence methods to generate CT-like images from magnetic resonance imaging (MRI), allowing for the geometric accuracy of CT needed for surgical planning while avoiding ionizing radiation [10]. A recent study by Michael et al. [11] showed that sCTs were able to measure anatomic structures of the spine with an accuracy within 1 mm of tCTs in 95% of children evaluated. An early feasibility cadaveric study suggested that these sCTs can safely and accurately support spine navigation [12]. In January of 2025, our institution performed the first MvIGS-navigated PSFs using MRI-generated sCTs in children. The purpose of this study is to report on the clinical and radiographic outcomes of this sCT cohort, describe the pedicle morphologies instrumented, compare the radiation exposure of sCT cases to tCT and FAFT cases, and finally report on any perioperative complications related to the use of sCT navigation. Materials and methods Preoperative protocol This study was approved by our institutional review board. Utilization of MvIGS navigation paired with tCT imaging became routine for one of the surgeons in 2023. For the remaining surgeons, their routine practice is to utilize the pure freehand or FAFT for pedicle screw instrumentation. At our institution, it is not the standard of care to obtain preoperative MRI of the spinal cord prior to PSFs unless there is an abnormality in the physical exam or the radiographic curve pattern (i.e. left thoracic curve, kyphosis, rapidly progressive deformity). Therefore, the implementation of sCTs required a change in our preoperative protocols. MRI was obtained using a 3.0T GE Signa Architect scanner (GE HealthCare Technologies Inc., Florence, SC, USA). The imaging protocol included sagittal T2-weighted (T2-WI) sequences, sagittal short tau inversion recovery (STIR) T2-WI, axial T2-WI, and a T1-weighted (T1-WI) 3D dual-echo gradient echo (GRE) sequence. The MRI acquisition was performed using a generic 3D RF-spoiled dual-echo GRE sequence with a total scan duration of 3 minutes. All GRE data sets were processed with BoneMRI® v1.8 software (MRIguidance B.V., Utrecht, The Netherlands) to generate sCTs; however this software was upgraded to v1.9 during this study. These nonsedated MRIs were performed no later than one day prior to the patient's surgery (to allow time for sCT conversion). Patient selection A retrospective review was performed of all patients who underwent primary, instrumented PSFs utilizing MvIGS navigation paired with sCT imaging between January 1, 2025 and December 31, 2025, yielding 20 patients. Given the novelty of this imaging modality, the first three sCT patients also received a tCT as a backup in case the sCT became unusable during the case. During the first case, prior to performing flash registration, the tCT data were loaded and a flash was performed, yielding 3,218 anatomic datapoints that were matched to the tCT. Then, without moving the patient or navigation machine, the sCT data were loaded and a flash was performed at the same vertebral levels, yielding 6,072 anatomic datapoints matched to the sCT (Fig. 1).Figure 1Comparison of anatomic datapoints (green dots) matched to a traditional CT (tCT) and to a synthetic CT (sCT) for the same patient, at the same vertebral registration levels during the index pediatric navigated case. CT, computed tomography. Comparison of anatomic datapoints (green dots) matched to a traditional CT (tCT) and to a synthetic CT (sCT) for the same patient, at the same vertebral registration levels during the index pediatric navigated case. CT, computed tomography. After the first 3 patients, the concern of sCT data corruption in the MvIGS machine seemed unfounded; therefore, the remaining patients only received an sCT. For the purposes of comparing radiation exposure to the patient, two matched cohorts were created. The sCT cohort was matched to a comparison cohort of 27 idiopathic scoliosis patients who received a preoperative tCT along with routine intraoperative fluoroscopy during their PSF between January 1, 2024 and December 31, 2024. All tCTs were performed using a standardized low-dose radiation protocol (Supplement 1). A third comparison cohort was created, matched by scoliosis etiology, including patients who received instrumented PSFs utilizing the FAFT during the same period as the sCT cases. Intraoperative protocol The implementation of MvIGS navigation utilizing sCT data did not change the routine intraoperative workflow for PSFs. During the early portion of exposure, according to the published best practice guidelines [13], the pedicle at the T12 vertebral level is marked with a radiopaque tool, and its position is confirmed using fluoroscopy (Fig. 2).Figure 2Intraoperative fluoroscopy imaging showing the pedicle at the T12 vertebral level marked with a radiopaque tool. Intraoperative fluoroscopy imaging showing the pedicle at the T12 vertebral level marked with a radiopaque tool. After this, fluoroscopy is not used again until the surgeon wants to check that the upper instrumented vertebra (UIV) and lowest instrumented vertebra (LIV) are horizontal/level and check successful rotational alignment after direct vertebral derotation activities. After exposure, wide facetectomies are performed, and then, depending on the spine’s flexibility, more extensive posterior column osteotomies are added if needed. Afterward, the reference frame for the MvIGS navigation system is attached to the spinous process of the LIV. A flash registration of the bottom 2-4 levels planned for instrumentation is performed. Specific anatomic points on the lamina of the 2-4 levels visualized in the flash registration are selected and correlated with the sCT imaging (VIDEO). Once this is confirmed, the surgeon checks in real time that the navigation is accurate by touching off on the cortical surfaces using a navigated awl or pedicle probe and confirming that the instrument is not floating off the bone on the navigation screen. Then pedicle screw pathways are created using either the navigated pedicle probe or the navigated drill. These pedicle screw pathways are probed primarily with a non-navigated ball-tipped probe. Then, with the assistance of the virtual trajectory that was saved (VIDEO), the pedicle screw is inserted with a non-navigated instrument on power. Datapoints collected Routine demographic variables were collected including age at the time of surgery, scoliosis diagnoses, sex, and time between MRI and surgery. Radiographic parameters collected included major curve magnitude, kyphosis, and lordosis at the preoperative and postoperative timepoints. Using true axial cuts of 20 sCTs, pedicle morphologies of all instrumented levels were classified as described by Sarwahi et al. [1]. Pedicles were categorized as follows: type A defined as pedicle diameters greater than 4 mm with a visible cancellous canal; type B between 2 and 4 mm with a visible cancellous canal; type C ≥ 2 mm with exclusively cortical channels only; and type D < 2 mm with exclusively cortical channels. To assess interobserver reliability, two independent reviewers (J.L. & E.O.) blinded to each other's assessments graded pedicle morphology in a subset of 5 patients (84 pedicles). Because pedicle morphology grades represent an ordinal scale (types A–D), quadratic weighted kappa was calculated rather than the intraclass correlation coefficient (ICC), which is reserved for continuous measurements. Raw agreement was 79.8% (67/84 pedicles), and the quadratic weighted kappa was 0.54, indicating moderate interobserver agreement by Landis and Koch criteria. Perioperative complications were collected. Preoperative radiation exposure from preoperative tCT imaging and intraoperative fluoroscopy was collected from the radiology department for the sCT, tCT, and FAFT cohorts. Radiation exposure was reported in milliGray (mGY). Statistical analysis For the sCT cohort, descriptive statistics were reported for the demographic and radiographic parameters. For the radiation exposure comparisons, a student's t-test was performed using SPSS Statistics for Mac, version 29 (SPSS Inc., Chicago, IL, USA). A P-value <.05 was considered statistically significant. Results Patient demographics A total of 20 patients underwent navigated PSFs using the MvIGS system paired with sCT data. The average age at surgery was 13.9 ± 3.7 years. Eleven patients were female. Most patients (n = 12, 60%) had juvenile or AIS diagnoses (Table 1). The mean preoperative major curve magnitude was of 61.2° ± 19.8°. The mean time between MRI acquisition and surgery was 48 ± 54 days.Table 1Demographics for 20 patients undergoing navigated sCT PSFs.Table 1Age at surgery13.9 ± 3.7 YearsSexMale9Female11DiagnosisJuvenile idiopathic3Adolescent idiopathic9Congenital2Neuromuscular/syndromic4Kyphoscoliosis2Cobb anglePreoperative61.2° ± 19.8°Postoperative18° ± 13.1°Percent correction72.5% ± 16.7%KyphosisPreoperative32.4° ± 17.9°Postoperative34.4° ± 8.6°LordosisPreoperative50° ± 16.0°Postoperative44.7° ± 12.7°Time between MRI and surgery48 ± 54 dayssCT, synthetic computed tomography; PSFs, posterior spinal fusions; MRI, magnetic resonance imaging. Table 1Demographics for 20 patients undergoing navigated sCT PSFs.Table 1Age at surgery13.9 ± 3.7 YearsSexMale9Female11DiagnosisJuvenile idiopathic3Adolescent idiopathic9Congenital2Neuromuscular/syndromic4Kyphoscoliosis2Cobb anglePreoperative61.2° ± 19.8°Postoperative18° ± 13.1°Percent correction72.5% ± 16.7%KyphosisPreoperative32.4° ± 17.9°Postoperative34.4° ± 8.6°LordosisPreoperative50° ± 16.0°Postoperative44.7° ± 12.7°Time between MRI and surgery48 ± 54 dayssCT, synthetic computed tomography; PSFs, posterior spinal fusions; MRI, magnetic resonance imaging. Demographics for 20 patients undergoing navigated sCT PSFs. sCT, synthetic computed tomography; PSFs, posterior spinal fusions; MRI, magnetic resonance imaging. Pedicle grading and screws A total of 335 pedicles were instrumented with type A pedicles accounting for the majority of pedicle morphologies (Table 2). Type C and Type D pedicles accounted for 9% and 4% of the pedicles instrumented, respectively, using a variety of screw diameters (Table 2).Table 2Pedicle morphologies instrumented in 20 patients undergoing navigated sCT PSFs.Table 2DiameterPedicle morphology (N = 20, 335 pedicles)Grand totalABCD519 (35.9%)21 (39.9%)11 (20.4%)3 (5.6%)545.554 (55.7%)19 (19.6%)16 (16.5%)8 (8.2%)97665 (82.3%)10 (12.7%)4 (5.1%)–796.565 (97.0%)1 (1.5%)1 (1.5%)–67738 (100%)–––38UIV, upper instrumented vertebra; LIV, lowest instrumented vertebra; sCT, synthetic computed tomography; PSFs, posterior spinal fusions. Table 2Pedicle morphologies instrumented in 20 patients undergoing navigated sCT PSFs.Table 2DiameterPedicle morphology (N = 20, 335 pedicles)Grand totalABCD519 (35.9%)21 (39.9%)11 (20.4%)3 (5.6%)545.554 (55.7%)19 (19.6%)16 (16.5%)8 (8.2%)97665 (82.3%)10 (12.7%)4 (5.1%)–796.565 (97.0%)1 (1.5%)1 (1.5%)–67738 (100%)–––38UIV, upper instrumented vertebra; LIV, lowest instrumented vertebra; sCT, synthetic computed tomography; PSFs, posterior spinal fusions. Pedicle morphologies instrumented in 20 patients undergoing navigated sCT PSFs. UIV, upper instrumented vertebra; LIV, lowest instrumented vertebra; sCT, synthetic computed tomography; PSFs, posterior spinal fusions. At the UIV, type A pedicles were most common and all pedicles at the LIV were type A. At the curve apex, type A pedicles again predominated, followed by type B pedicles (Table 3).Table 3Location of instrumented pedicles in 20 patients undergoing navigated sCT PSFs.Table 3LocationPedicle morphology (N = 20, 335 pedicles)ABCDUIV22 (58%)10 (26%)5 (13%)1 (3%)LIV38 (100%)0 (0%)0 (0%)0 (0%)Apex18 (58%)8 (26%)2 (6%)3 (10%)Total243 (72%)52 (15%)32 (9%)12 (4%)sCT, synthetic computed tomography; PSFs, posterior spinal fusions; UIV, upper instrumented vertebra; LIV, lowest instrumented vertebra. Table 3Location of instrumented pedicles in 20 patients undergoing navigated sCT PSFs.Table 3LocationPedicle morphology (N = 20, 335 pedicles)ABCDUIV22 (58%)10 (26%)5 (13%)1 (3%)LIV38 (100%)0 (0%)0 (0%)0 (0%)Apex18 (58%)8 (26%)2 (6%)3 (10%)Total243 (72%)52 (15%)32 (9%)12 (4%)sCT, synthetic computed tomography; PSFs, posterior spinal fusions; UIV, upper instrumented vertebra; LIV, lowest instrumented vertebra. Location of instrumented pedicles in 20 patients undergoing navigated sCT PSFs. sCT, synthetic computed tomography; PSFs, posterior spinal fusions; UIV, upper instrumented vertebra; LIV, lowest instrumented vertebra. Radiation exposure The mean radiation exposure to the 27 patients receiving a preoperative tCT scan, combined with routine intraoperative fluoroscopy, was 5.44 mGy (Table 4). Given that the sCT modality did not expose the patient to any radiation, the only radiation exposure was due to routine intraoperative fluoroscopy, reflecting an 83.5% reduction in radiation (P = .001). In comparison, during the same time period, 3 surgeons utilized FAFT to insert pedicle screws during PSFs, producing a mean radiation exposure of 6.96 mGy. Comparing the mean sCT case radiation exposure of 0.90 mGy to the FAFT cases, sCT cases resulted in an 87.1% reduction in radiation exposure (P < .001).Table 4Radiation exposure for 20 sCT patients compared to a matched cohort of FAFT patients.Table 4Radiation exposure based on pedicle screw placement techniqueExposure sourcesCTN = 20tCTN = 27FAFTN = 135P-valueTraditional CT mGy–3.72 ± 0.92 (3.2-7.2)––FluromGy0.90 ± 1.37 (0.12-5.1)1.72 ± 3.94 (0.15-20.73)6.96 ± 8.21 (0.69-56.62)<.001secs6.65 ± 7.16 (0.70-26.00)10.14 ± 14.95 (1.40-59.50)58.34 ± 34.06 (5.89-180.10)<.001Total mGy0.90 ± 1.37 (0.12-5.1)5.44 ± 4.28 (3.35-25.43)∗6.96 ± 8.21 (0.69-56.62)<.001FAFT cohort: Total mGy:Surgeon 1 (n = 38): 5.05 (0.93-33.78).Surgeon 2 (n = 48): 4.65 (0.69-31.8).Surgeon 3 (n = 49): 10.78 (2.48-56.62).∗Radiation exposure from backup CT for the first 3 patients was not included in the comparison.sCT, synthetic computed tomography; FAFT, fluoroscopy-assisted freehand technique; tCT, traditional computed tomography. Table 4Radiation exposure for 20 sCT patients compared to a matched cohort of FAFT patients.Table 4Radiation exposure based on pedicle screw placement techniqueExposure sourcesCTN = 20tCTN = 27FAFTN = 135P-valueTraditional CT mGy–3.72 ± 0.92 (3.2-7.2)––FluromGy0.90 ± 1.37 (0.12-5.1)1.72 ± 3.94 (0.15-20.73)6.96 ± 8.21 (0.69-56.62)<.001secs6.65 ± 7.16 (0.70-26.00)10.14 ± 14.95 (1.40-59.50)58.34 ± 34.06 (5.89-180.10)<.001Total mGy0.90 ± 1.37 (0.12-5.1)5.44 ± 4.28 (3.35-25.43)∗6.96 ± 8.21 (0.69-56.62)<.001FAFT cohort: Total mGy:Surgeon 1 (n = 38): 5.05 (0.93-33.78).Surgeon 2 (n = 48): 4.65 (0.69-31.8).Surgeon 3 (n = 49): 10.78 (2.48-56.62).∗Radiation exposure from backup CT for the first 3 patients was not included in the comparison.sCT, synthetic computed tomography; FAFT, fluoroscopy-assisted freehand technique; tCT, traditional computed tomography. Radiation exposure for 20 sCT patients compared to a matched cohort of FAFT patients. ∗Radiation exposure from backup CT for the first 3 patients was not included in the comparison. sCT, synthetic computed tomography; FAFT, fluoroscopy-assisted freehand technique; tCT, traditional computed tomography. Complications In the sCT cohort, one patient experienced a transcranial motor evoked potential signal change during intraoperative traction, which resolved after traction was removed and was unrelated to pedicle screw placement. There were no other IOMN changes during or after pedicle screw instrumentation. There were no cerebral spinal fluid leak events. There were no postoperative neurologic deficits noted. In only 1 instance was sCT pedicle screw navigation planned but not performed, due to corrupted imaging that resulted in the absence of clearly visualized vertebral body anatomy including some posterior element anatomy. This patient thus received pedicle screw instrumentation using a FAFT (see Fig. 3).Figure 3Corrupted synthetic CT imaging showing the absence of visualized vertebral body anatomy. CT, computed tomography. Corrupted synthetic CT imaging showing the absence of visualized vertebral body anatomy. CT, computed tomography. Discussion The use of MvIGS navigation to aid in the precise instrumentation of pedicle screws is not novel and has been proven accurate both in the idiopathic and neuromuscular populations [14,15]. Traditionally this technology required preoperative tCT imaging to aid in pedicle screw navigation which spared only the surgeon and surgical team from ionizing radiation. However, Simony et al. [3] showed a 4.3-fold relative risk of cancer development in AIS patients at 25 years of follow-up, highlighting the importance of reducing the ionizing radiation burden to our patients too. In January 2025, we performed the first navigated pediatric case in the USA using MvIGS navigation paired with sCT imaging. The purpose of this study was to report on the clinical, radiographic, and complication outcomes of this cohort, describing the pedicle morphologies instrumented and comparing the radiation exposure of sCT cases to tCT and FAFT cases. sCT imaging was successfully used to navigate pedicle screw instrumentation in idiopathic, neuromuscular, syndromic, and congenital scoliosis patients. There were no IOMN changes that occurred during pedicle screw instrumentation, and no patients had postoperative neurologic deficits. Using sCTs paired with MvIGS navigation, difficult type C and D pedicles were instrumented. For radiation exposure, sCT cases combined with routine intraoperative fluoroscopy produced 83.5% less radiation than the preoperative tCT scan needed for MvIGS navigation. In comparison, for a matched cohort of patients who received pedicle screws via the FAFT, sCT cases produced 87.1% less radiation exposure. The potential medical applications for sCTs are numerous [16]; however, to date, most orthopaedic applications have been related to using sCTs for diagnostic purposes such as detecting osteoarthritis [17] of the knee or evaluating sacroiliac [18] and hip anatomy [19]. The only comparable study in the literature showing the real-world application of using sCTs in spinal deformity surgeries was done by Kashtiara et al. [20]. In their study, they used sCT imaging in a robotic navigation system, finding that it was as accurate to the tCT scan imaging they routinely used, with a 98% reduction in radiation. The mean age of their patient population was 55 ± 19 years, and most had degenerative scoliosis. Given the significant workflow differences between robotic and MvIGS navigation, few additional comparisons can be made. In our study, the authors utilized MvIGS navigation with tCT imaging for 2.5 years prior to attempting an sCT case. Unsurprisingly, the MvIGS navigation system could visualize the surface anatomy in children with kyphosis or scoliosis secondary to idiopathic, syndromic, neuromuscular, and congenital diagnoses. For the authors of this study, the integration of sCTs for cognitively impaired, nonambulatory patients with neuromuscular scoliosis has been more challenging since this population often requires sedation with intubation to decrease motion during their MRI. Because of this, the authors still utilize a tCT scan (on the morning of surgery) for this population, potentially highlighting some disparities in the application of sCT technology. Fortunately, Sakhrekar et al. [15] have shown excellent outcomes in neuromuscular scoliosis (NMS) patients utilizing tCT scans paired with MvIGS navigation. In our cohort of 20 children undergoing navigated sCT spine deformity surgeries, we experienced no cerebrospinal fluid leaks or IOMN events related to pedicle screw instrumentation. In addition, no patients experienced postoperative neurologic deficits; however, the absence of neurologic events in this small, underpowered cohort should not be interpreted as definitive evidence of safety. Importantly, intraoperative and postoperative tCT imaging were not obtained to confirm final pedicle screw position as this was not even part of our workflow when utilizing MvIGS navigation paired with tCT scans in prior years; accordingly, formal accuracy assessment cannot be made from this series. Minor asymptomatic breaches cannot be excluded. Instead, after pedicle screws were inserted, and after both rods were successfully reduced, long-cassette orthogonal radiographs were obtained. These intraoperative radiographs were the final check of pedicle screw position and overall deformity correction before final tightening of set caps and decortication was performed. While tCT is considered the gold standard for postoperative screw assessment, Kim et al. [21] demonstrated that applying three standardized radiographic criteria to intraoperative radiographs yielded sensitivity of 0.87-0.94 and accuracy of 0.96-0.98 compared with tCT. In addition, Sapkas et al. [22] found no statistically significant difference between plain radiographs and tCT in detecting malposition in the absence of neurologic deficit. Importantly, the clinical significance of minor cortical breaches detectable only on CT remains questionable—the Scoliosis Research Society (SRS)/Pediatric Orthopaedic Society of North America (POSNA) systematic review reported a 94.9% accuracy rate across over 13,000 pediatric pedicle screws with exceedingly rare neurologic complications [23]. So, while this study was not designed to evaluate the accuracy of sCT navigation in placing pedicle screws, the lack of complications in this early cohort combined with our intraoperative radiographic evaluation of the instrumentation and deformity correction has yielded promising outcomes. The greatest advantage of replacing tCT imaging for sCTs in MvIGS navigation cases is the drastic reduction of radiation exposure for the patient. As mentioned previously, there is a 4.3-fold increased risk of cancers experienced by AIS patients over a lifetime, which is well above the cancer risk of matched patients without AIS. The multiple episodes of ionizing radiation exposure during the monitoring and treatment of these patients cannot be ignored. We were able to show an 83.5% reduction in radiation exposure between tCT (5.44 mGy) and sCT (0.90 mGy) navigated cases, with most of the radiation exposure in sCT cases coming from the minimal use of fluoroscopy to confirm the location of T12 along with a leveled UIV and LIV after rod insertion. We were surprised however, at the significant difference in radiation exposure (87.1%) between FAFT and sCT cases. There is no doubt that the FAFT is safe and the gold-standard method of instrumenting pedicle screws. Anecdotally, the authors will also admit that FAFT cases take less time than most navigated cases. However, our study showed that both the surgeon and patient were exposed to significantly more radiation during FAFT cases (6.96 mGy) than during sCT cases (0.90 mGy). Interestingly, in comparison to the patients receiving a traditional preoperative CT scan (5.44 mGy), FAFT cases (6.96 mGy) still resulted in significantly more radiation exposure, to the surgeon and the patient. The literature supports that the cumulative radiation burden borne by surgically treated children with scoliosis is substantial and disproportionately driven by intraoperative and perioperative imaging. Intraoperative fluoroscopy alone accounts for approximately 78% of the annual radiation exposure in operatively managed AIS patients, yielding exposure rates 8- to 14-fold higher than those observed in nonoperatively treated cohorts [24]. When navigation or robotic-assisted techniques necessitate a preoperative tCT, the perioperative effective dose may increase by as much as 25-fold compared with FAFT, with a single tCT contributing 12-20 mSv to the patient's cumulative burden [25]. By contrast, surveillance and postoperative radiography contribute comparatively modest doses, with each full-spine radiograph delivering approximately 0.05-0.08 mSv and cumulative radiographic exposure over a typical treatment course totaling less than 1 mSv [26]. The estimated lifetime cancer induction risk for surgically treated pediatric patients is approximately 1 in 222, compared with 1 in 1,418 for nonsurgical patients [2]. These data underscore the clinical imperative for navigation technologies that meaningfully reducing ionizing radiation exposure to the developing pediatric patient. This study has several limitations. The sCT cohort size is relatively small with a wide range of diagnoses as it represents a single-center early experience of a single surgeon. This may limit the generalizability of the findings. Radiation comparisons may have also been influenced by differences in surgeon technique, deformity severity, and the lack of randomization between the FAFT group and sCT-based navigation cohorts. Radiation exposure to surgeons and operating room staff was not directly measured using wearable dosimeters, limiting the evaluation of the full occupational impact between the cohorts. Postoperative tCT imaging was not obtained; this limits the ability to formally assess pedicle screw accuracy as minor asymptomatic breaches cannot be excluded. This series is not powered to draw conclusions regarding complication rates; the absence of major complications in 20 patients should not be interpreted as established safety. Complication data were not collected for the tCT and FAFT comparison cohorts, precluding cross-cohort complication comparison. The FAFT comparison cohort reflects three specific surgeons’ fluoroscopy practices; fluoroscopy use varies widely by surgeon and institution, and freehand techniques with only confirmatory imaging would likely produce lower radiation exposure than reported here. sCT image quality was assessed subjectively by the operating surgeon; formal quantitative quality metrics were not applied. Despite this fact, the imaging quality was deemed sufficient to successfully navigate pedicle screws in all but one patient whose imaging was shown in Fig. 2. In addition, while raw agreement was 79.8%, the interobserver reliability for pedicle morphology yielded only moderate agreement with a quadratic weighted kappa of 0.54. The cancer risk estimates cited by Simony et al. [3] reflect prebiplanar radiology (aka EOS) imaging practices and may overestimate current risk in patients managed with modern low-dose protocols. Finally, the cost of procuring a machine-vision navigation system paired with MRI-derived sCT software was not evaluated. Future studies should include larger prospective series with standardized protocols, postoperative tCT-verified accuracy assessment, and multisurgeon, multicenter designs to further evaluate the reproducibility and long-term outcomes of sCT-based navigation in pediatric spinal deformity surgery. In conclusion, this proof-of-concept feasibility series demonstrates that MRI-derived sCT-based navigation using a machine-vision navigation system is a promising approach for posterior spinal fusion in pediatric deformity patients. Integration of sCTs resulted in a substantial reduction in ionizing radiation during both preoperative and intraoperative imaging compared with both FAFT and tCT-based navigation while maintaining screw placement across a range of pedicle morphologies. These findings support the potential for a near radiation-free navigation approach to pediatric spine surgery and represent an early step toward minimizing cumulative radiation exposure in this vulnerable population. Larger prospective studies with formal accuracy verification are needed before definitive conclusions regarding safety and efficacy can be drawn. Author contributions Jaysson T. Brooks: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Project administration, Methodology, Investigation, Conceptualization. Judy-Mae Lima: Writing – review & editing, Writing – original draft, Investigation, Formal analysis, Data curation. Emmanuel Ogu: Writing – review & editing, Writing – original draft, Investigation, Data curation. David Thornberg: Writing – review & editing, Project administration, Investigation, Formal analysis, Data curation. Daniel J. Sucato: Writing – review & editing, Writing – original draft. Amy McIntosh: Writing – review & editing, Writing – original draft. Megan Johnson: Writing – review & editing, Writing – original draft. Brandon Ramo: Writing – review & editing, Writing – original draft. Vivek Kalia: Writing – review & editing, Writing – original draft, Validation. J. Herman Kan: Writing – review & editing, Writing – original draft, Validation. Ethics approval and consent Complete written informed consent was obtained from the patient, their guardian or legal representative for the publication of this study and accompanying images. Declaration of the use of generative AI and AI-assisted technologies in the writing process During the preparation of this work, the author(s) used OpenEvidence® for review and summary of the current literature. The authors reviewed and edited the output as needed and take full responsibility for the content of the published article. Funding No funding sources were used for this study. Declaration of competing interests The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Jaysson Brooks reports a relationship with OrthoPediatrics Corp and Orthofix Medical Inc that includes consulting or advisory and a relationship with Highridge Medical LLC that includes speaking and lecture fees. Amy McIntosh reports a relationship with Globus Medical Inc and Medtronic that includes consulting or advisory. Megan Johnson reports a relationship with Globus Medical Inc and DePuy Synthes that includes consulting or advisory. Daniel Sucato receives royalties from Globus Medical. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.