85. Comparison of novel machine vision image guidance system with a 3D fluoroscopy-based platform for pedicle screw placement: a randomized prospective study of radiation, accuracy, and cost

Abstract

<h3>BACKGROUND CONTEXT</h3> The use of image guidance and robotics platforms has increased patient safety, accuracy and operative efficiency and reduced revision rates during pedicle screw placement procedures. However, traditional intraoperative 3D or computational tomography (CT) imaging produces potentially harmful ionizing radiation and increases operative time to register the patient. An image guidance platform using machine vision through high resolution stereoscopic cameras and structured visible light does not use ionizing radiation while it builds a 3D topographical map of the patient's anatomy for navigation use. <h3>PURPOSE</h3> We aimed to compare a novel machine vision image guidance system (MvIGS) system to a widely used 3D fluoroscopy (3D) platform by comparing radiation exposure, and pedicle screw accuracy. <h3>STUDY DESIGN/SETTING</h3> A randomized prospective comparative cohort study of consecutive patients undergoing posterior lumbar instrumented fusion. <h3>PATIENT SAMPLE</h3> Adults diagnosed with a spinal condition requiring surgical treatment and planning to have posterior open lumbar fusion with pedicle screws implanted into 1-4 vertebral levels. <h3>OUTCOME MEASURES</h3> Outcome measures included: mean intraoperative fluoroscopy time and dose, mean preoperative CT dose length product (DLP) and mean day 2 postoperative CT DLP, pedicle screw accuracy and rate of revision surgery. <h3>METHODS</h3> Patients were randomized 1:1 to MvIGS or 3D and underwent posterior lumbar instrumented fusion. Radiation doses were recorded from pre- and postoperative CT and intraoperative fluoroscopy. Independent blinded radiologists reviewed pedicle screw accuracy. <h3>RESULTS</h3> A total of 303 (154 MvIGS, 149 3D) pedicle screws were placed in 63 patients (33 MvIGS, 30 3D) over the 12-month study period. Mean age and indication for surgery were similar between both groups. Mean intraoperative fluoroscopy time and doses were significantly reduced in MvIGS compared to 3D (5.2±4.1s vs 79.6±23.3s, p<0.001 and 133.0±215.8cGycm2 vs 3704.1±1640.5cGycm2, p<0.001, respectively). This represents a relative reduction of 93.5% in the total intraoperative time and 96.4% reduction in the total dose. Mean preoperative CT DLP and mean day 2 postoperative CT DLP were significantly reduced in MvIGS compared to 3D (590.8±389.6mGy*cm vs 1008.9±616.3mGy*cm, p<0.001 and 518.1±205.6mGy*cmvs 958.7±449.6mGy*cm, p<0.001, respectively). This represents a relative reduction of 41.4% and 46.0% relative reduction in the preoperative CT dose and postoperative total DLP, respectively. The MvIGS group required an average of 1.2 registrations each case with an average of 2447 (SD=961.3) data points registered with mean registration time of 106s (SD=52.1s). A rapid re-registration mechanism was utilized in 22% (n=7/33) of cases. Pedicle screw accuracy was high in both groups MvIGS (98.7%) vs 3D (97.3%) with no pedicle breach >2mm in either group (p>0.001). No neurovascular injuries occurred, and no patients needed return to theatre for screw repositioning. <h3>CONCLUSIONS</h3> Both MvIGS and 3D image guidance platforms demonstrated high accuracy for pedicle screw placement. MvIGS demonstrated significant reduction of 93.5% and 96.4% in intraoperative radiation time and dose. MvIGS may provide similar accuracy to contemporary image guidance platforms without requiring 3D fluoroscopy or radiolucent operating tables. Reducing registration times and the need for specialized equipment may significantly reduce operating room costs. <h3>FDA DEVICE/DRUG STATUS</h3> This abstract does not discuss or include any applicable devices or drugs.