A Novel Drill Navigation Template Combines Preoperative Simulation in Expansive Open-Door Laminoplasty

Automatic Pavlov ratio measurement method based on spinal landmarks identification by a deep-learning model

BACKGROUND

Cervical canal stenosis is one of the important pathogenic factors of cervical spondylosis. The accuracy of the Pavlov ratio measurement is crucial for the diagnosis and treatment of cervical spinal stenosis. Manual measurement is influenced by observer variability, accompanied by its inefficiency, which affects clinical evaluation.

PURPOSE

To automatically and accurately measure the Pavlov ratio, we develop a novel deep-learning model by detecting keypoints of cervical spine and measure the Pavlov ratio on plain lateral cervical spine radiographs.

METHODS

We developed a two-stage deep-learning model; in the first stage, we employ the YOLOX model as the object detection network to locate the ROIs containing the vertebral bodies and spinous processes; in the second stage, we introduce the high-resolution net (HRNet) as keypoint detection network and a series of deconvolutional networks (DNs) as the heatmap-based regressor. Based on the mentioned combining algorithms, we can rapidly detect the 38 keypoints in plain lateral cervical spine radiographs, and then measure the Pavlov ratio of the cervical spine. Radiographs from Shanghai Changhai Hospital (a total of 874) were split into training and test subsets (787 and 87 radiographs, respectively). One hundred twelve cases from Shanghai Changzheng Hospital and 108 cases from Shanghai Fourth People's Hospital are used as external validation datasets.

RESULTS

Our proposed model successfully achieved the objective of automating the recognition of spinal landmarks with the mean absolute error (MAE)ranged from 0.05 to 0.08, and the symmetric mean absolute percentage error (SMAPE) ranged from 4.54% to 6.43%. The achieved accuracy is comparable to that of seasoned medical professionals and notably surpasses the performance of junior physicians (SMAPE ranged from 8.74% to 26.19%). Furthermore, our model demonstrated excellent accuracy in external validation experiments (SMAPE ranged from 4.40% to 5.95%).

CONCLUSION

This study presents a novel YOLOX-HRNet-DN model to assist landmarks identification on lateral cervical spine radiographs and demonstrates excellent accuracy in measuring the Pavlov ratio. The proposed method could provide a potential tool for the automatic estimation of the Pavlov ratio to improve the efficiency and accuracy of the treatment workflow.

Accuracy of the trough position in expansive open-door cervical laminoplasty using computer navigation techniques: a single-centre retrospective study

PURPOSE

This study aims to assess whether computer navigation can improve the accuracy of the trough position and clinical outcomes of expansive open-door cervical laminoplasty (EOLP).

METHODS

We reviewed a single centre of 28 conventional EOLP and 24 computer navigation EOLP cases. The conventional group had 102 laminae while the navigation group had 88. The distance from the medial cortex to the pedicle on the open-door side (OD) and hinge side (HD) was measured. Furthermore, the area of the spinal canal corresponding to each lamina before and after the surgical procedure was also measured. We then compared the differences in radiographic parameters and clinical outcomes between the two groups.

RESULTS

OD and HD were smaller in the navigation group compared to the conventional group, and the enlarged area of the spinal canal was larger in the navigation group than in the conventional group. The Japanese Orthopaedic Association (JOA) scores one year after the surgical procedure improved in both groups compared to the pre-operative period, and the JOA recovery rate was higher in the navigation group. The incidence of hinge fracture was lower in the navigation group, and the incidence of C5 palsy and axial pain was not statistically different between the two groups.

CONCLUSION

The use of computer navigation techniques has the potential to significantly improve the accuracy of EOLP compared to conventional procedures. It has been shown to more fully expand the spinal canal and contribute to clinical efficacy.

A Novel Method of Making Hinges Using a Newly Designed Sharp Rongeur to Enhance Radiological and Clinical Outcomes in French-Door Cervical Expansive Laminoplasty

OBJECTIVE

Although the lamina open angle of making hinges is closely related to the outcomes of French-door laminoplasty (FDL) for treatment of cervical spondylosis, there have been no methods to predict the lamina open angle preoperatively as yet. The aim of this study was to investigate the accuracy of predicting the laminal open angle using our newly designed sharp rongeur, and to compare the postoperative outcomes and complications between the methods of making hinges using the newly designed sharp rongeur and the traditional high-speed micro-drill during the FDL.

METHODS

This was a single-center retrospective study. Following the approval of the institutional ethics committee, a total of 39 patients (Male: 28; Female: 11) diagnosed with cervical spondylos who underwent FDL in our institution between January 2018 and May 2019 were enrolled. Patients were divided into two groups based on the method of making hinges (sharp rongeur: 22 cases; high-speed micro-drill: 17 cases). The average age at surgery was 59.1 years (range: 16-85 years). The radiological parameters, clinical outcomes, modified Japanese Orthopaedic Association (mJOA) scale score, and the recovery rate of mJOA were recorded and compared between the groups, respectively. The radiological parameters and clinical measurements at pre- and post-operation stages were compared using the paired-sample t-test, the Wilcoxon signed-rank test, and the Friedman's test, and variables in the two groups were analyzed using an unpaired Student's t-test or a Mann-Whitney U test.

RESULTS

The average follow-up period was 20.4 months (range: 14.0-25.9 months), the postoperative open angle was 60.13° ± 3.69° in the rongeur group with 22.78° ± 4.34° of angular enlargement, which was significantly lower than that of 68.96° ± 1.00° in the micro-drill group with 32.75° ± 4.22° of angular enlargement (U = 19.000, p < 0.001). The rongeur group showed a higher fusion rate (34.1% vs 14.7%, χ²  = 11.340, p = 0.001), and a lower fracture rate of the lamina (7.8% vs 25.5%, χ²  = 14.185, p < 0.001) at 1-month post-surgery, compared to the micro-drill group. There were no significant differences in the clinical outcomes and postoperative complications between the two groups (p > 0.05), except in the recovery rate of mJOA scores (0.836 ± 0.138 vs 0.724 ± 0.180, U = 115.000, p = 0.042) and neck disability index (NDI) at the final follow-up (7.55 ± 10.65 vs 14.71 ± 8.72, U = 94.000, p = 0.008).

CONCLUSIONS

The special sharp rongeur with a tip angle of 20° could be a preferred method to make hinges during FDL, which can predict the laminal open angle accurately and enlarge it to about 23°, thus reducing the fracture rate and accelerating the bony fusion of hinges compared with the outcomes of the traditional micro-drill method.

The Development of Novel 2-in-1 Patient-Specific, 3D-Printed Laminectomy Guides with Integrated Pedicle Screw Drill Guides

OBJECTIVE

To determine if 2-in-1 patient-specific laminectomy and drill guides can be safely used to perform laminectomy and pedicle screw insertion.

METHODS

This was a cadaveric study designed to test novel 2-in-1 patient-specific laminectomy guides, with modular removable pedicle screw drill guides. Three-dimensional (3D) printing has not been applied to laminectomy. This cadaveric study tests novel 2-in-1 patient-specific laminectomy guides, with modular removable pedicle screw drill guides. Computed tomography (CT) scans of 3 lumbar spines were imported into 3D Slicer. Spinal models and patient-specific guides were created and 3D printed. The bones were cleaned to visualize and record the under surface of the lamina during laminectomy. Pedicle screws and laminectomies were performed with the aid of patient-specific guides. CT scans were performed to compare planned and actual screw and laminectomy positions.

RESULTS

Thirty screws were inserted in 15 lumbar vertebrae by using the integrated 2-in-1 patient-specific drill guides. There were no cortical breaches on direct examination, or on postoperative CT. Digital video analysis revealed the burr tip did not pass deep to the inner table margin of the lamina in any of the 30 laminectomy cuts. Average surgical time was 4 minutes and 46 seconds (standard deviation, 1 min 38 sec).

CONCLUSIONS

This study has explored the development of novel 2-in-1 patient-specific, 3D-printed laminectomy guides with integrated pedicle screw drill guides, which are accurate and safe in the laboratory setting. These instruments have the potential to simplify complex surgical steps, and improve accuracy, time, and cost.

3D printed navigation template-guided minimally invasive percutaneous plate osteosynthesis for distal femoral fracture: A retrospective cohort study

OBJECTIVE

To evaluate functional and radiological outcomes in distal femoral fracture patients (distal femoral shaft fractures or metaphyseal fractures) receiving indirect reduction and internal fixation with the minimally invasive percutaneous plate osteosynthesis (MIPPO) assisted by 3D printing navigation templates.

METHODS

This retrospective study included all adult (≥18 years) patients who underwent MIPPO for distal femoral fracture (AO32/33) at PLA 960 Hospital, Jinan, China between January 1, 2013 and December 31, 2016. 3D printed navigation templates were used to assist surgery. Functional assessment was performed using the Knee Society Score (KSS) and the Functional Knee Society Score (FKSS). Postoperative deformity was assessed with 3D-CT reconstruction using contralateral (non-fracture) side as the reference.

RESULTS

A total of 54 patients underwent MIPPO for distal femoral fracture during the review period (34 and 20 for conventional and template-guided MIPPO, respectively). At 10-14 months, the KSS score was 160.6 ± 18.1 and 167.5 ± 17.2 in the conventional and template-guided MIPPO groups, respectively (p = 0.18). The FKSS was 77.6 ± 9.4 in the conventional MIPPO group vs. 81.0 ± 8.5 in the template-guided MIPPO group (p = 0.15). Femoral length difference was smaller in template-guided MIPPO group (3.31 ± 1.53 vs. 9.50 ± 4.49 mm in the control; p < 0.01). Template-guided MIPPO group also had smaller anatomic lateral distal femoral angle (aLDFA) difference (1.57° ± 0.72° vs. 3.89° ± 1.72° in the control; p < 0.01), anatomic posterior distal femoral angle (aPDFA) difference (1.95° ± 0.78° vs. 5.04° ± 1.78°; p < 0.01), and anteversion angle difference (2.52° ± 1.00° vs. 8.23° ± 4.07°; p < 0.01). The rate of length discrepancy (>20 mm), coronal angulation deformity (aLDFA difference >5°), sagittal angulation deformity (aPDFA difference >10°), and rotational malalignment (anteversion angle >10°) was 5.9%, 20.5%, 2.9%, and 32.4% in the control group. No patients in the template-guided MIPPO group had deformity of any type.

CONCLUSIONS

MIPPO assisted by 3D printing navigation templates could practically improve surgical accuracy and eliminate postoperative rotational deformity in patients with distal femoral fractures.