W‐Transformer: Accurate Cobb angles estimation by using a transformer‐based hybrid structure

Abstract

Background: Scoliosis is a type of spinal deformity, which is harmful to a person's health. In severe cases, it can trigger paralysis or death. The measurement of Cobb angle plays an essential role in assessing the severity of scoliosis. Purpose: The aim of this paper is to propose an automatic system for landmark detection and Cobb angle estimation, which can effectively help clinicians diagnose and treat scoliosis. Methods: A novel hybrid framework was proposed to measure Cobb angle precisely for clinical diagnosis, which was referred as W-Transformer due to its wshaped architecture. First, a convolutional neural network of cascade residual blocks as our backbone was designed. Then a transformer was fused to learn the dependency information between spine and landmarks. In addition, a reinforcement branch was designed to improve the overlap of landmarks, and an improved prediction module was proposed to fine-tune the final coordinates of landmarks in Cobb angles estimation. Besides, the public Accurate Automated Spinal Curvature Estimation (AASCE) MICCAI 2019 challenge was served as data set. It supplies 609 manually labeled spine anterior-posterior (AP) X-ray images, each of which contains a total of 68 landmark labels and three Cobb Angles tags. Results: From the perspective of the AASCE MICCAI 2019 challenge, we achieved a lower symmetric mean absolute percentage error (SMAPE) of 8.26% for all Cobb angles and the lowest averaged detection error of 50.89 in terms of landmark detection, compared with many state-of -the-art methods. We also provided the SMAPEs for the Cobb angles of the proximal-thoracic (PT), the main-thoracic (MT), and the thoracic-lumbar (TL) area, which are 5.27%, 14.59%, and 20.97% respectively, however, these data were not covered in most previous studies. Statistical analysis demonstrates that our model has obtained a high level of Pearson correlation coefficient of 0.9398 (p < 0.001), which shows excellent reliability of our model. Our model can yield 0.9489 (p < 0.001), 0.8817 (p < 0.001), and 0.9149 (p < 0.001) for PT, MT, and TL, respectively. The overall variability of Cobb angle measurement is less than 4 • , implying clinical value. And the mean absolute deviation (standard deviation) for three regions is 3.64 • (4.13 • ), 3.84 • (4.66 • ), and 3.80 • (4.19 • ). The results of Student paired t-test indicate that no statistically significant differences are observed between manual measurement and our automatic approach (p-value is always >0.05). Regarding the diagnosis of scoliosis (Cobb angle >10 • ), the proposed method achieves a high sensitivity of 0.9577 and a specificity of 0.8475 for all spinal regions. Conclusions: This study offers a brand-new automatic approach that is potentially of great benefit of the complex task of landmark detection and Cobb angle 3246

The timing of tether breakage influences clinical results after VBT

“…The main finding of this work is that a tether breakage occurring in the first 12 months after VBT leads to a significant loss of correction. When the breakage occurs after 12 months or does not occur, the mean loss of correction is limited and less than 4°, which would imply a clinically relevant change [16].…”

Comparative Accuracy of Three Artificial Intelligence Algorithms for Automatic Cobb Angle Measurement in Scoliosis: An Analysis Using Disease-specific Teaching Data

“…It can also measure CA with high accuracy in postures other than standing, such as supine and lateral bending. There are algorithms for measuring curves other than the main curve [14,17,18], but all of them reported only on AIS cases and there was no mention of supine or lateral bending postures. The measurement of CA in the minor curves and in the supine or lateral bending position is necessary to classify the cases based on classi cations, such as the Lenke classi cation, and to determine the range of xation in surgery.…”

“…However, these methods require the manual selection of the appropriate end vertebrae and endplates by an operator. In this context, there has been a recent marked increase in reports on arti cial intelligence (AI) algorithms for fully automated measurement of CA using convolutional neural networks (CNN) [14][15][16][17][18][19][20][21][22][23][24][25][26][27]. The measurement error of the CA in these reports ranged from 1.9° to 9.9°, which is similar to or lower than that in the manual and computer-assisted manual methods.…”

“…Most of these reports focused only on AIS, and only few have mentioned ASD cases [22,26]. Moreover, reports on the measurement of minor curves are limited [14,17,18], and only one report has veri ed the accuracy in the supine or lateral bending posture [27]. These problems, which include a lack of versatility for AIS and ASD, lack of measurement of minor curves, and lack of proven accuracy in various postures, remain unresolved.…”

Automatic Lenke classification of adolescent idiopathic scoliosis with deep learning

“…Totally, it can be divided into two categories: segmentation‐based estimation methods 8 , 9 , 10 and landmark‐based estimation methods. 11 , 12 , 13 , 14 , 15 , 16 The segmentation‐based estimation methods primarily involve the initial segmentation of either the entire spinal region or each individual vertebra. 9 Specifically, the procedure involved extracting the region of interest (ROI) of the spine, outlining the spine's edges, and subsequently applying U‐Net for spine segmentation to calculate its curvature.…”