Machine learning for automated 3-dimensional segmentation of the spine and suggested placement of pedicle screws based on intraoperative cone-beam computer tomography

OBJECTIVEThe goal of this study was to develop and validate a system for automatic segmentation of the spine, pedicle identification, and screw path suggestion for use with an intraoperative 3D surgical navigation system.METHODSCone-beam CT (CBCT) images of the spines of 21 cadavers were obtained. An automated model-based approach was used for segmentation. Using machine learning methodology, the algorithm was trained and validated on the image data sets. For measuring accuracy, surface area errors of the automatic segmentation were compared to the manually outlined reference surface on CBCT. To further test both technical and clinical accuracy, the algorithm was applied to a set of 20 clinical cases. The authors evaluated the system’s accuracy in pedicle identification by measuring the distance between the user-defined midpoint of each pedicle and the automatically segmented midpoint. Finally, 2 independent surgeons performed a qualitative evaluation of the segmentation to judge whether it was adequate to guide surgical navigation and whether it would have resulted in a clinically acceptable pedicle screw placement.RESULTSThe clinically relevant pedicle identification and automatic pedicle screw planning accuracy was 86.1%. By excluding patients with severe spinal deformities (i.e., Cobb angle > 75° and severe spinal degeneration) and previous surgeries, a success rate of 95.4% was achieved. The mean time (± SD) for automatic segmentation and screw planning in 5 vertebrae was 11 ± 4 seconds.CONCLUSIONSThe technology investigated has the potential to aid surgeons in navigational planning and improve surgical navigation workflow while maintaining patient safety.

Download Full-text

Workflow and performance of intraoperative CT, cone-beam CT, and robotic cone-beam CT for spinal navigation in 503 consecutive patients

Neurosurgical FOCUS ◽

10.3171/2021.10.focus21467 ◽

2022 ◽

Vol 52 (1) ◽

pp. E7

Keyword(s):

Image Quality ◽

Pedicle Screw ◽

Posterior Instrumentation ◽

Cone Beam Ct ◽

Demographic Data ◽

Pedicle Screws ◽

Cone Beam ◽

Intraoperative Ct ◽

Spinal Navigation ◽

And Performance

OBJECTIVE A direct comparison of intraoperative CT (iCT), cone-beam CT (CBCT), and robotic cone-beam CT (rCBCT) has been necessary to identify the ideal imaging solution for each individual user’s need. Herein, the authors sought to analyze workflow, handling, and performance of iCT, CBCT, and rCBCT imaging for navigated pedicle screw instrumentation across the entire spine performed within the same surgical environment by the same group of surgeons. METHODS Between 2014 and 2018, 503 consecutive patients received 2673 navigated pedicle screws using iCT (n = 1219), CBCT (n = 646), or rCBCT (n = 808) imaging during the first 24 months after the acquisition of each modality. Clinical and demographic data, workflow, handling, and screw assessment and accuracy were analyzed. RESULTS Intraoperative CT showed image quality and workflow advantages for cervicothoracic cases, obese patients, and long-segment instrumentation, whereas CBCT and rCBCT offered independent handling, around-the-clock availability, and the option of performing 2D fluoroscopy. All modalities permitted reliable intraoperative screw assessment. Navigated screw revision was possible with each modality and yielded final accuracy rates > 92% in all groups (iCT 96.2% vs CBCT 92.3%, p < 0.001) without a difference in the accuracy of cervical pedicle screw placement or the rate of secondary screw revision surgeries. CONCLUSIONS Continuous training and an individual setup of iCT, CBCT, and rCBCT has been shown to permit safe and precise navigated posterior instrumentation across the entire spine with reliable screw assessment and the option of immediate revision. The perceived higher image quality and larger scan area of iCT should be weighed against the around-the-clock availability of CBCT and rCBCT technology with the option of single-handed robotic image acquisition.

Download Full-text

Intraoperative cone beam CT in hybrid operating room set-up as an alternative to postoperative CT for pedicle screw breach detection

10.21203/rs.3.rs-256928/v1 ◽

2021 ◽

Author(s):

Simon Peh ◽

Julian Pfarr ◽

Jost Philipp Schäfer ◽

Jan-Hendrik Christensen ◽

Anindita Chatterjea ◽

...

Keyword(s):

Radiation Dose ◽

Pedicle Screw ◽

Operating Room ◽

Cone Beam Ct ◽

Pedicle Screws ◽

Roc Curves ◽

Cone Beam ◽

Screw Placement ◽

Hybrid Operating Room ◽

Effective Radiation Dose

Abstract Background CT is considered the gold standard for detecting pedicle breach. However, CBCT may be a viable and low radiation dose alternative, to provide intraoperative feedback to surgeons to permit in-room revisions of misplaced screws Methods To assess the ability and reliability of intraoperative cone-beam CT (CBCT) from a robotic C-arm in a hybrid operating room (OR) two hundred forty-one pedicle screws were inserted in cervical, thoracic and lumbar spine of 7 cadavers, followed by CBCT and CT imaging. The CT images served as the standard of reference. Agreement on screw placement between both imaging systems was assessed using Cohen’s Kappa coefficient (κ). Sensitivity, Specificity, Receiver operating characteristic (ROC), area under the empirical and fitted ROC curves (AUC) were computed to assess CBCT as a diagnostic tool compared to CT. The patient effective radiation dose (ED) was calculated for comparison. A systematic literature review was performed to provide perspective to the obtained results. Results Almost perfect agreement in assessing pedicle screw grading between CBCT and CT was observed (κ = 0.84). The sensitivity and specificity of CBCT were 0.84 and 0.98, respectively. The AUC derived from the empirical and fitted ROC curves were 0.95 and 0.96, respectively. Conclusion Intraoperative CBCT by C-arm in a hybrid OR is highly reliable in identification of screw placement at significant dose reduction.

Download Full-text

Image-Guided Spinal Surgery and Robotics in MIS: Where Are We Now?

Neurologico Spinale Medico Chirurgico ◽

10.15562/nsmc.v1i2.80 ◽

2018 ◽

Vol 1 (2) ◽

pp. 2

Author(s):

Chiung Chyi Shen

Keyword(s):

Pedicle Screw ◽

Spinal Surgery ◽

Surgical Navigation ◽

Improve Patient Safety ◽

Intraoperative Imaging ◽

Spinal Instrumentation ◽

Pedicle Screws ◽

Computer Assisted ◽

Navigation Systems ◽

Image Guided

Use of pedicle screws is widespread in spinal surgery for degenerative, traumatic, and oncological diseases. The conventional technique is based on the recognition of anatomic landmarks, preparation and palpation of cortices of the pedicle under control of an intraoperative C-arm (iC-arm) fluoroscopy. With these conventional methods, the median pedicle screw accuracy ranges from 86.7% to 93.8%, even if perforation rates range from 21.1% to 39.8%.The development of novel intraoperative navigational techniques, commonly referred to as image-guided surgery (IGS), provide simultaneous and multiplanar views of spinal anatomy. IGS technology can increase the accuracy of spinal instrumentation procedures and improve patient safety. These systems, such as fluoroscopy-based image guidance ("virtual fluoroscopy") and computed tomography (CT)-based computer-guidance systems, have sensibly minimized risk of pedicle screw misplacement, with overall perforation rates ranging from between 14.3% and 9.3%, respectively."Virtual fluoroscopy" allows simultaneous two-dimensional (2D) guidance in multiple planes, but does not provide any axial images; quality of images is directly dependent on the resolution of the acquired fluoroscopic projections. Furthermore, computer-assisted surgical navigation systems decrease the reliance on intraoperative imaging, thus reducing the use of intraprocedure ionizing radiation. The major limitation of this technique is related to the variation of the position of the patient from the preoperative CT scan, usually obtained before surgery in a supine position, and the operative position (prone). The next technological evolution is the use of an intraoperative CT (iCT) scan, which would allow us to solve the position-dependent changes, granting a higher accuracy in the navigation system.

Download Full-text

Radiation doses of sliding gantry ct-based as compared to mobile cone-beam CT-based navigated pedicle screw placement in a homogenous cohort

Brain and Spine ◽

10.1016/j.bas.2021.100273 ◽

2021 ◽

Vol 1 ◽

pp. 100273

Author(s):

Sebastian Ille ◽

Lea Baumgart ◽

Bernhard Meyer ◽

Sandro Krieg

Keyword(s):

Pedicle Screw ◽

Cone Beam Ct ◽

Cone Beam ◽

Screw Placement ◽

Radiation Doses ◽

Pedicle Screw Placement

Download Full-text

A Novel Augmented-Reality-Based Surgical Navigation System for Spine Surgery in a Hybrid Operating Room: Design, Workflow, and Clinical Applications

Operative Neurosurgery ◽

10.1093/ons/opz236 ◽

2019 ◽

Vol 18 (5) ◽

pp. 496-502 ◽

Cited By ~ 18

Author(s):

Erik Edström ◽

Gustav Burström ◽

Rami Nachabe ◽

Paul Gerdhem ◽

Adrian Elmi Terander

Keyword(s):

Augmented Reality ◽

Radiation Exposure ◽

Pedicle Screw ◽

Operating Room ◽

Wound Closure ◽

Surgical Navigation ◽

Pedicle Screws ◽

Screw Placement ◽

Hybrid Operating Room ◽

Pedicle Screw Placement

Abstract BACKGROUND Treatment of several spine disorders requires placement of pedicle screws. Detailed 3-dimensional (3D) anatomic information facilitates this process and improves accuracy. OBJECTIVE To present a workflow for a novel augmented-reality-based surgical navigation (ARSN) system installed in a hybrid operating room for anatomy visualization and instrument guidance during pedicle screw placement. METHODS The workflow includes surgical exposure, imaging, automatic creation of a 3D model, and pedicle screw path planning for instrument guidance during surgery as well as the actual screw placement, spinal fixation, and wound closure and intraoperative verification of the treatment results. Special focus was given to process integration and minimization of overhead time. Efforts were made to manage staff radiation exposure avoiding the need for lead aprons. Time was kept throughout the procedure and subdivided to reflect key steps. The navigation workflow was validated in a trial with 20 cases requiring pedicle screw placement (13/20 scoliosis). RESULTS Navigated interventions were performed with a median total time of 379 min per procedure (range 232-548 min for 4-24 implanted pedicle screws). The total procedure time was subdivided into surgical exposure (28%), cone beam computed tomography imaging and 3D segmentation (2%), software planning (6%), navigated surgery for screw placement (17%) and non-navigated instrumentation, wound closure, etc (47%). CONCLUSION Intraoperative imaging and preparation for surgical navigation totaled 8% of the surgical time. Consequently, ARSN can routinely be used to perform highly accurate surgery potentially decreasing the risk for complications and revision surgery while minimizing radiation exposure to the staff.

Download Full-text