Precise Cobb Angle Measurement System Based on Spinal Images Merging Function

The Lenke classification system is widely utilized as the preoperative evaluation protocol for adolescent idiopathic scoliosis (AIS). However, manual measurement is susceptible to observer-induced variability, which consequently impacts the evaluation of progression. The goal of this investigation was to develop an automated Lenke classification system utilizing innovative deep learning algorithms. Using the database from the First Affiliated Hospital of Sun Yat-sen University, the whole spinal x-rays images were retrospectively collected. Specifically, images collection was divided into AIS and control group. The control group consisted of individuals who underwent routine health checks and did not have scoliosis. Afterwards, relative features of all images were annotated. Deep learning was implemented through the utilization of the key-point based detection method to realize the vertebral detection, and Cobb angle measurement and scoliosis classification were performed based on relevant standards. Besides, the segmentation method was employed to achieve the recognition of lumbar vertebral pedicle to determine the type of lumbar spine modifier. Finally, the model performance was further quantitatively analyzed. In the study, a total of 2082 spinal x-ray images were collected from 407 AIS patients and 227 individuals in the control group. The model for vertebral detection achieved an F1-score of 0.809 for curve type evaluation and an F1-score of 0.901 for thoracic sagittal profile. The intraclass correlation efficient (ICC) of the Cobb angle measurement was 0.925. In the analysis of performance for vertebra pedicle segmentation model, the F1-score of lumbar modification profile was 0.942, the intersection over union (IOU) of the target pixels was 0.827, and the Hausdorff distance (HD) was 6.565 ± 2.583 mm. Specifically, the F1-score for ultimate Lenke type classifier was 0.885. This study has constructed an automated Lenke classification system by employing the deep learning networks to achieve the recognition pattern and feature extraction. Our models require further validation in additional cases in the future.

X-ray images of human spine have the characteristics of low contrast and much noise generally. Due to the variable shape of the spine edge, the traditional spine image segmentation method requires a lot of preprocessing and can not automatically perform image segmentation, which is inefficient and cannot meet the clinical work requirements well. In order to solve the above problems, this paper use deep learning algorithms to identify the X-ray images of spine, and a cobb angle measurement method for scoliosis based on U-Net model for X-ray images is proposed. Improve U-Net model to get more image features, thereby improving recognition accuracy. Experiments prove that the improved U-Net model can accurately segment the vertebrae in the spine X-ray image with high accuracy and speed.

Intra- and inter-observer reliability of Cobb angle measurements in patients with adolescent idiopathic scoliosis.

Introduction Scoliosis is characterized by an abnormal curvature of the spine in the coronal plane. Idiopathic scoliosis is the most prevalent type, though specific causes are sometimes identifiable. Genetic factors significantly influence adolescent idiopathic scoliosis (AIS), which is diagnosed through clinical and radiographic evaluations, primarily using the Cobb angle to measure curvature severity. The classification of scoliosis severity ranges from mild scoliosis, where sometimes the absence of pain is encountered, to moderate and severe, which is usually associated with lancinating pain. Early onset and high progression rates in idiopathic scoliosis are indicative of poorer prognoses. Methods The study analyzed 197 radiographic images from a private clinic database between December 2023 and April 2024. Inclusion criteria focused on anteroposterior images of the thorax and abdomen, excluding unclear and non-spinal images. Manual Cobb angle measurements were performed using RadiAnt DICOM Viewer 2020.2, followed by automated measurements using the Cobb Angle Calculator software. Discrepancies led to further image processing with enhanced color contrast for improved visualization. Data were analyzed using GraphPad InStat to assess error margins between manual and automated measurements. Results Initial results indicated discrepancies between manual and automated Cobb angle measurements. Enhanced image processing improved accuracy, demonstrating the efficacy of both manual and automated techniques in evaluating spinal deformities. Statistical analysis revealed significant error margins, prompting a refined approach for minimizing measurement errors. Discussion The study highlights the importance of accurate Cobb angle measurement in diagnosing and classifying scoliosis. Manual measurements, while reliable, are time-consuming and prone to human error. Automated methods, particularly those enhanced by machine learning algorithms, offer promising accuracy and efficiency. The integration of image processing techniques further enhances the reliability of scoliosis evaluation. Conclusion Accurate assessment of scoliosis through Cobb angle measurement is crucial for effective diagnosis and treatment planning. The study demonstrates that combining manual techniques with advanced automated methods and image processing significantly improves measurement accuracy. Such an approach is intended to support better clinical outcomes. Future research should focus on refining these technologies for broader clinical applications.

Open-source artificial intelligence models (OSAIM) find free applications in various industries, including information technology and medicine. Their clinical potential, especially in supporting diagnosis and therapy, is the subject of increasingly intensive research. Due to the growing interest in artificial intelligence (AI) for diagnostic purposes, we conducted a study evaluating the capabilities of AI models, including ChatGPT and Microsoft Bing, in the diagnosis of single-curve scoliosis based on posturographic radiological images. Two independent neurosurgeons assessed the degree of spinal deformation, selecting 23 cases of severe single-curve scoliosis. Each posturographic image was separately implemented onto each of the mentioned platforms using a set of formulated questions, starting from ‘What do you see in the image?’ and ending with a request to determine the Cobb angle. In the responses, we focused on how these AI models identify and interpret spinal deformations and how accurately they recognize the direction and type of scoliosis as well as vertebral rotation. The Intraclass Correlation Coefficient (ICC) with a ‘two-way’ model was used to assess the consistency of Cobb angle measurements, and its confidence intervals were determined using the F test. Differences in Cobb angle measurements between human assessments and the AI ChatGPT model were analyzed using metrics such as RMSEA, MSE, MPE, MAE, RMSLE, and MAPE, allowing for a comprehensive assessment of AI model performance from various statistical perspectives. The ChatGPT model achieved 100% effectiveness in detecting scoliosis in X-ray images, while the Bing model did not detect any scoliosis. However, ChatGPT had limited effectiveness (43.5%) in assessing Cobb angles, showing significant inaccuracy and discrepancy compared to human assessments. This model also had limited accuracy in determining the direction of spinal curvature, classifying the type of scoliosis, and detecting vertebral rotation. Overall, although ChatGPT demonstrated potential in detecting scoliosis, its abilities in assessing Cobb angles and other parameters were limited and inconsistent with expert assessments. These results underscore the need for comprehensive improvement of AI algorithms, including broader training with diverse X-ray images and advanced image processing techniques, before they can be considered as auxiliary in diagnosing scoliosis by specialists.

Development of a computer-aided Cobb measurement method and evaluation of its reliability. To reduce the variability of Cobb angle measurement by developing the computer-aided method and to investigate if the developed method is sensitive to observer skill levels or experiences. Therapeutic decisions for scoliosis heavily rely on the Cobb angle measured from consecutive radiographs. The manual Cobb measurement is subject to human errors. The observer error is 3 to 10 degrees resulted from different end-vertebrae selection and/or manually drawing variable best-fit lines to the endplates of the end-vertebrae. A fussy Hough transform technique was used to develop a computer-aided method to detect the vertebral endplates. The Cobb angle, upper end-vertebra, and lower end-vertebra were then measured automatically. The computer-aided method was tested twice by each of 3 observers in 84 posteroanterior radiographs from patients with adolescent idiopathic scoliosis. The intraobserver and interobserver errors were analyzed. Both the intraobserver and interobserver reliability analyses resulted in the intraclass correlation coefficients higher than 0.9 for the Cobb angle. The average intraobserver and interobserver errors were less than 3 degree for the Cobb angle, and less than 0.3 levels for both the upper and lower end-vertebral identification. There were no significant differences in the measurement variability between groups of curve location (thoracic, thoracolumbar, and lumbar), curve direction (right and left), curve magnitude (curves less than 25 degree, between 25 and 45 degrees, and more than 45 degree), and observer experience (experienced observer and inexperienced observers). Compared with the documented results, variability of the Cobb measurement is reduced by using the developed computer-aided method. This method can help orthopedic surgeons measure the Cobb angle more reliably during scoliosis clinics.

Objective: To develop a clinically feasible apparatus for measuring and marking patient’s gravity line position on spinal x-ray images. Background: Spinal balance of patients with deformity or instability is an important clinical parameter for patient diagnosis, surgery planning and treatment outcome evaluation. Traditional assessment of spinal balance is performed on full spine x-ray images using C7 plumb line relative to central sacral line in the coronal plane or posterior superior corner of S1 in the sagittal plane. Researchers have developed measurement techniques using a force plate synchronized with x-ray equipment and offline x-ray image processing. The complexity of these techniques limits their clinical applications. In this study, we present a new system to mark gravity line directly on an x-ray image. Methods: The system consists of a force plate, a signal conditioner and processor, a motion controller and a motor assembly that moves a radio-opaque marker. The processor calculates the location of gravity line and marker position, and transmits motion commands to the motion controller, which drives a motor assembly and moves the marker to appropriate positions. The acquired x-ray image has a mark indicating the instantaneous gravity line position. Evaluation: A prototype was installed in the radiology department. More than 200 patients were instructed to stand on the force plate. Exposure was performed as in routine x-ray imaging. The system does not affect routine x-ray workflow and provides valuable information for spine surgeons.

The purpose of this study was to assess the impact of patient position on the magnitude of the coronal Cobb angle measurements in relation to the change of position using plain radiograph on non-ambulatory children with myelodysplasia. Whole-spine radiographs with the patient sitting generally are preferred for the diagnosis and monitoring of progression of scoliosis in neuromuscular patients. Supine, supine traction, and sitting push-up positions have been used as substitutes, although there is no general consensus validating if these positions correlate with the sitting position. The magnitude of the Cobb angles in neuromuscular scoliosis may vary greatly depending on the position of the patient. Radiographs of 39 myelodysplastic, non-ambulatory children were evaluated to assess the impact of change in positions (unsupported sitting, sitting push-up, supine, and supine traction) on coronal Cobb angle measurement using plain whole-spine radiographs. The mean difference in thoracic Cobb angle measurements between sitting and all other positions ranged from 6° to 12°. At the lumbar level, the Cobb angles ranged from 12° to 16°. Statistically significant differences in the Cobb angle measurements were identified between plain radiographs of the whole spine with the patient in the unsupported sitting position compared to sitting push-up, supine, and supine traction positions. The data support that the magnitude of the Cobb angles in neuromuscular scoliosis varies greatly depending on the position of the patient. III.

Ultrasensitive and stable X-ray detection using zero-dimensional lead-free perovskites

In order to determine the reliability of the Cobb angle measurement as it is used in the clinical management of scoliosis, a methodological survey was carried out. In the measurement of a Cobb angle two phases can be distinguished: (a) the production of a spinal radiograph and (b) the measurement of the angle itself. In respect of the first phase, the variation in production of the radiographs was calculated on Cobb angle measurements made by one investigator on serial radiographs of patients who underwent spinal fusion for scoliosis and therefore had a fixed spinal curvature. For the second phase, the accuracy of Cobb angle measurement was investigated by comparing measurements on the same radiographs of 46 scoliosis patients obtained by three investigators, namely two orthopaedic surgeons and an orthopaedic fellow who was assigned to a school screening project. Results were expressed as a Spearman correlation coefficient and a standard deviation of the differences. The Spearman correlation coefficient was 0.98 for the repeated radiographs (production variation) and also 0.98 for the repeated measurements on one radiograph (interobserver measurement variation). The standard deviation of the differences in Cobb angle for the repeated radiographs amounted to 3.2 degrees and for the repeated measurements on one radiograph it was 2.0 degrees. Although there is a good reproducibility of the Cobb angle measurement between different investigators, the variation in production of a spinal radiograph is an important source of error. This should be taken into account when making decisions in scoliosis management.

Growing rods in Early Onset Scoliosis: The current scenario

Accurate segmentation of vertebral blocks in X-ray images of the whole spine is necessary for intelligent diagnosis of spinal diseases. However, the vertebral block in X-ray spine image has similar appearance with background, which cause a huge challenge to segmentation performance. Even though the classical Unet network and existing popular methods have been proved as effective approaches in medical image segmentation, they still create some conglutination between vertebral blocks in spine X-ray images. To overcome this problem, we propose a novel spine X-ray image segmentation method based on transformer and adaptive optimized postprocessing to improve segmentation performance. Firstly, we adopt the Segmenter network based on Transformer framework to obtain initial result. Then, an adaptive post-processing optimization strategy is proposed to optimize segmentation results. Finally, experiments on AASCE 2019 challenge dataset show that our model can gain best performance compared with existing methods.

Lung function asymmetry in children with congenital and infantile scoliosis

CdTe photon-counting detectors can offer great advantages over conventional X-ray detectors but, at the same time, require specific pre-processing to attain artifact free final images. A complete pre-processing procedure for one of such detectors is detailed in this chapter.

In current computed tomography (CT) architecture, both X-ray tubes and X-ray detectors are rotated mechanically around an object to collect a sufficient number of projections. This architecture has been shown to not be fast enough for patients with high or irregular heart rates. Furthermore, both X-ray beams and detectors of the current architecture are made wide enough, so that the entire object is covered in the lateral direction without data truncation. Although novel acquisition protocols have recently been developed to reduce a radiation exposure, the high radiation dose from CT imaging remains a heightened public concern (especially for cardiac CT). The current CT architecture is a major bottleneck to further increase the temporal resolution and reduce the radiation dose. To overcome these problems, we present an innovative stationary-sources rotating-detectors CT (SSRD-CT) architecture based on the three stationary distributed X-ray sources and three smaller rotating X-ray detectors. Each distributed X-ray source has ~ 100 distinctive X-ray focal spots, and each detector has a narrower width compared with the conventional CT detectors. The SSRD-CT will have a field-of-view of 200 mm in diameter at isocenter, which is large enough to image many internal organs, including hearts. X-rays from the distributed sources are activated electronically to simulate the mechanical spinning of conventional single-beam X-ray sources with a high speed. The activation of individual X-ray beam will be synchronized to the corresponding rotating detector at the opposite end. Three source-detector chains can work in parallel to acquire three projections simultaneously and improve temporal resolution. Lower full-body radiation dose is expected for the proposed SSRD-CT because X-rays are restricted to irradiate a local smaller region. Taken together, the proposed SSRD-CT architecture will enable ≤50-ms temporal resolution and reduce radiation dose significantly.