The study introduces a novel method for automatic segmentation of vertebral column tissue from MRI images. The paper describes a method that combines multiple stages of Machine Learning techniques to recognize and separate different tissues of the human spine. For the needs of this paper, 50 MRI examinations presenting lumbosacral spine of patients with low back pain were selected. After the initial filtration, automatic vertebrae recognition using Cascade Classifier takes place. Afterwards the main segmentation process using the patch based Active Appearance Model is performed. Obtained results are interpolated using centripetal Catmull–Rom splines. The method was tested on previously unseen vertebrae images segmented manually by 5 physicians. A test validating algorithm convergence per iteration was performed and the Intraclass Correlation Coefficient was calculated. Additionally, the 10-fold cross-validation analysis has been done. Presented method proved to be comparable to the physicians (FF = 90.19 ± 1.01%). Moreover results confirmed a proper algorithm convergence. Automatically segmented area correlated well with manual segmentation for single measurements (falser¯=normal0.8336) and for average measurements (falser¯=normal0.9068) with p = 0.05. The 10-fold cross-validation analysis (FF = 91.37 ± 1.13%) confirmed a good model generalization resulting in practical performance.
IntroductionRecently the EOS imaging system (EOS Imaging, Paris, France) has provided advancements in 3D spinal modeling. Advancements include low radiation as well as fast and accurate reconstructed measurements of spinal parameters. There is a paucity of studies analyzing the reproducibility of the EOS Imaging System and the sterEOS software in the production of 3D spinal models for children with adolescent idiopathic scoliosis (AIS). Objectives The purposes of the study were 1) to determine the intraclass correlation (ICC) for both the inter-observer and intra-observer in the measurements of Cobb angles in AP view as well as the Cobb angles in the lateral view; 2) to assess the ICC for inter-and intra-observer in the axial vertebral rotation (AVR) of the apex vertebra; 3) to compare differences of spinal parameters between two examiners and two trials; 4) to determine how long a 3D reconstruction of the spine takes. Methods Bilateral x-ray images of fifteen patients (age: 6 -15 years old, 5 males, 10 females) were retrospectively selected. These EOS images were uploaded into the sterEOS computer program. Within the software, spinal and pelvic parameters were identified manually to construct a 3D model of the spine. The sterEOS software calculates the Cobb angles, angles of lordosis, angles of kyphosis, and the AVRs of the apex vertebra. The 3D modeling was performed independently by two examiners. Each examiner modeled each patient's spine in two spaced out trials. The ICC between inter-and intra-observers were calculated and compared statistically. Results and discussionBoth the inter-and intra-observers showed excellent reproducibility for the Cobb angles in the proximal segment (ICC: 0.72 -0.91), kyphosis (ICC: 0.85-0.92), and lordosis (ICC: 0.82 -0.95). No significant differences were found between angle differences (0.35°to 2.4°). In contrast to the traditional radiography, the sterEOS provides a better high quality view within the sagittal plane. A moderate inter-observer ICC for the Cobb angle in the distal segment (ICC = 0.67) indicates the examiners have to carefully adjust the alignment and vertebrae in 3D rather than in 2D following the automatic computation from the EOS software. The interobserver ICC for the AVR in the lumbar region (0.80) is higher than the thoracic or thoracolumbar region (0.65), but with high differences of AVR (4.0°-6.3°). The average time that two examiners spent per subject ranged from 34.6 to 37.4 minutes. Conclusion and significance EOS provides significantly reliable and accurate spinal modeling in the measurement of children with AIS. Exposure to less radiation as compared to other radiographic modality allows EOS to offer acceptable quality view of the spine in the sagittal and transversal plane. sagittal balance and predictive equations to determine lumbopelvic compensatory patterns (LPCP). These equations are used to guide surgical decision making and technique selection. Although other lumbopelvic compensation equations are available, these have not been compared wi...
Morphological analysis of the scoliotic spine is based on two-dimensional X-rays: coronal and sagittal. The three-dimensional character of scoliosis has raised the necessity for analyzing scoliosis in three planes. We proposed a new user-friendly method of graphical presentation of the spine in the third plane-the Spine Axial Presentation (SAP). Eighty-five vertebrae of patients with scoliosis were analyzed. Due to different positions during X-rays (standing) and computer tomography (CT) (supine), the corresponding measurements cannot be directly compared.As a solution, a software creating Digital Reconstructed Radiographs (DRRs) from CT scans was developed to replace regular X-rays with DRRs. Based on the measurements performed on DRRs, the coordinates of vertebral bodies central points were defined. Next, the geometrical centers of vertebral bodies were determined on CT scans. The reproducibility of measurements was tested with Intraclass Correlation Coefficient (ICC), using p = 0.05. The intra-observer reproducibility and inter-observer reliability for vertebral body central point's coordinates ( x, y, z) were high for results obtained based on DRRs and CT scans, as well as for comparison results obtained based on DRR and CT scans. Based on two standard radiographs, it is possible to localize vertebral bodies in 3D space. The position of vertebral bodies can be present in the Spine Axial Presentation.
Introduction: Several techniques for pedicle screw placement have been described including freehand techniques, fluoroscopy assisted, computed tomography (CT) guidance, and robotics. Image-guided surgery offers the potential to combine the benefits of CT guidance without the added radiation. This study investigated the ability of a neural network to place lumbar pedicle screws with the correct length, diameter, and angulation autonomously within radiographs without the need for human involvement. Materials and Methods: The neural network was trained using a machine learning process. The method combines the previously reported autonomous spine segmentation solution with a landmark localization solution. The pedicle screw placement was evaluated using the Zdichavsky, Ravi, and Gertzbein grading systems. Results: In total, the program placed 208 pedicle screws between the L1 and S1 spinal levels. Of the 208 placed pedicle screws, 208 (100%) had a Zdichavsky Score 1A, 206 (99.0%) of all screws were Ravi Grade 1, and Gertzbein Grade A indicating no breech. The final two screws (1.0%) had a Ravi score of 2 (<2 mm breech) and a Gertzbein grade of B (<2 mm breech). Conclusion: The results of this experiment can be combined with an image-guided platform to provide an efficient and highly effective method of placing pedicle screws during spinal stabilization surgery.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.