Automated Cobb Angle Measurement for Adolescent Idiopathic Scoliosis Using Convolutional Neural Network

Caesarendra, Wahyu; Rahmaniar, Wahyu; Mathew, John; Thien, Ady

doi:10.3390/diagnostics12020396

Cited by 30 publications

(18 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…When the literature is examined, it can be seen that research works focused on specific solutions for scoliosis. Studies such as [47,56,57,68,73,75] are remarkable as they aimed Cobb angle measurement and comparisons with human-side values. Even, the study [65] performed just angle measurements.…”

Section: Discussionmentioning

confidence: 99%

Development of a CapsNet and Fuzzy Logic Decision Support System for Diagnosing the Scoliosis and Planning Treatments via Schroth Method

Goral

Köse

2022

IEEE Access

View full text Add to dashboard Cite

Scoliosis is a disease caused by the spine curving. It is treatable but physiotherapists may do different measurements for curvature angles. That's a problem affecting the treatment planning. This study aims to develop a Deep Learning-based decision support system, which diagnoses scoliosis and plans treatments via Schroth method. The system has an interpretable and explainable CapsNet model processing x-ray image to detect 68-point vertebrae and make Cobb angle measurements. By using angle values and patient parameters, treatment is planned through an automated Schroth definition and the Fuzzy Logic. In the evaluations, the CapsNet had dominating findings (some of them are MSE: 0.0038, PCC: 0.93, Accuracy: 0.98). The Fuzzy Logic model was accurate at exercise plans for past cases. Also, physiotherapists and patients had positive feedback for the system usage, trustworthiness, diagnosis, treatment planning and tracking. As a conclusion, the system ensures advancements for automated diagnosis and treatment of scoliosis.

show abstract

Section: Discussionmentioning

confidence: 99%

Development of a CapsNet and Fuzzy Logic Decision Support System for Diagnosing the Scoliosis and Planning Treatments via Schroth Method

Goral

Köse

2022

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Participants of both groups received spinal full-length anteroposterior and lateral radiographs (Philips TH-VS Dr) to determine compliance with the inclusion criteria ( Macagno and O’Brien, 2006 ; Enad et al, 2008 ) prior to their participation in the study ( Figure 1 ). The angle between the parallel line of the upper-end plate of T10 and the parallel line of the lower-end plate of L2 is Cobbs’ angle ( Caesarendra et al, 2022 ). The Cobbs’ angle of the thoracolumbar segment on each patient’s lateral radiographs was measured using Surgimap software (Nemaris, Inc.).…”

Section: Methodsmentioning

confidence: 99%

Impacts of gait biomechanics of patients with thoracolumbar kyphosis secondary to Scheuermann’s disease

Cheng

Jiang

Chen

et al. 2023

Front. Bioeng. Biotechnol.

View full text Add to dashboard Cite

Introduction: Thoracolumbar kyphosis (TLK) is a common feature in patients with spinal deformities. However, due to limited studies, the impacts of TLK on gait have not been reported. The objective of the study was to quantify and evaluate the impacts of gait biomechanics of patients with TLK secondary to Scheuermann’s disease.Methods: Twenty cases of Scheuermann’s disease patients with TLK and twenty cases of asymptomatic participants were recruited into this study. And the gait motion analysis was conducted.Results: The stride length was shorter in the TLK group compared to control group (1.24 ± 0.11 m vs. 1.36 ± 0.21 m, p = 0.04). Compared to control group, the stride time and step time were more prolonged in the TLK group (1.18 ± 0.11s vs. 1.11 ± 0.08 s, p = 0.03; 0.59 ± 0.06 s vs. 0.56 ± 0.04 s, p = 0.04). The gait speed of the TLK group was significantly slower than it of control group (1.05 ± 0.12 m/s vs. 1.17 ± 0.14 m/s, p = 0.01); In the sagittal plane, the range of motion (ROM) of the hip in the TLK group was significantly smaller than that of the control group (37.71 ± 4.35° vs. 40.05 ± 3.71°, p = 0.00). In the transverse plane, the adduction/abduction ROMs of the knee and ankle, as well as the internal and external rotation of the knee, were smaller in TLK group than ROMs in the control group (4.66 ± 2.21° vs. 5.61 ± 1.82°, p = 0.00; 11.48 ± 3.97° vs. 13.16 ± 5.6°, p = 0.02; 9.00 ± 5.14° vs. 12.95 ± 5.78°, p = 0.00).Discussion: The main finding of this study was that measurements of gait patterns and joint movement of the TLK group were significantly lower than those of the control group. And these impacts have the potential to exacerbate degenerative progress of joints in the lower extremities. These abnormal features of gait can also serve as a guideline for physicians to focus on TLK in these patients.

show abstract

“…Previous angle assessment models focused on identifying the four corners of each vertebra and using them as landmarks for measuring the Cobb angle or formed bounding boxes around each vertebra. [5,6]. Unlike these previous models, ALEXS generates complete masks for each vertebra, fully outlining its shape.…”

Section: Innovations Innovation 1 -Adaptive Loss Engine For X-ray Seg...mentioning

confidence: 99%

Adaptive loss engine for x-ray segmentation (ALEXS) for scoliosis intervention: assess digital segmentation and angle approximation

Shao,

2024

Medical Imaging 2024: Image-Guided Procedures, Robotic Interventions, and Modeling

View full text Add to dashboard Cite

Estimating the severity of scoliosis is time consuming and imprecise. Doctors currently manually identify each vertebra and measure the Cobb angles, a measurement used for scoliosis. This paper aims to contribute to developing a fully automated method of estimating Cobb angles. Historically, the greatest challenge of this goal has been identifying the location of vertebrae within the x-ray image This paper proposes the use of an image segmentation model for this purpose, since they can identify objects within images with high accuracy. However, their results are less accurate when given xrays. To solve this issue, a new specialized model was trained on additional data composed entirely of x-rays, and it was named Adaptive Loss Engine for X-Ray Segmentation (ALEXS). ALEXS is a self-improving model capable of automatically identifying vertebrae, providing noticeable improved segmentation results compared to models that have not been trained on x-rays. One method of helping the model identify more vertebrae is altering the original x-ray image without changing the locations of the vertebrae. It was found that among many image processing techniques, sharpening the image and increasing its contrast had the largest positive effect on the results, allowing ALEXS to identify many more vertebrae than before. Based on the results that were obtained, using ALEXS combined with altered images produces superior results compared to some previous attempts, with improvements in the accuracy of the produced segments. These improved methods allow for a more accurate end-to-end process for automatically diagnosing scoliosis.

show abstract

Automated Cobb Angle Measurement for Adolescent Idiopathic Scoliosis Using Convolutional Neural Network

Cited by 30 publications

References 22 publications

Development of a CapsNet and Fuzzy Logic Decision Support System for Diagnosing the Scoliosis and Planning Treatments via Schroth Method

Development of a CapsNet and Fuzzy Logic Decision Support System for Diagnosing the Scoliosis and Planning Treatments via Schroth Method

Impacts of gait biomechanics of patients with thoracolumbar kyphosis secondary to Scheuermann’s disease

Adaptive loss engine for x-ray segmentation (ALEXS) for scoliosis intervention: assess digital segmentation and angle approximation

Contact Info

Product

Resources

About