Direct Estimation of Spinal Cobb Angles by Structured Multi-output Regression

Sun, Haoliang; Zhen, Xiantong; Bailey, Chris; Rasoulinejad, Parham; Yin, Yilong; Li, Shuo

doi:10.1007/978-3-319-59050-9_42

Cited by 60 publications

(50 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The challenging automated analysis of the Cobb angle describing the severity of a scoliotic curve has been confronted with various approaches, ranging from non‐ML methods such as the fuzzy Hough transform to deep learning techniques. Sun et al used a regression SVM to predict the Cobb angle from coronal radiographs, with a very good accuracy (relative root mean squared error of 21.6%) highlighting a potential clinical use. Zhang et al trained a deep ANN to predict the vertebral slopes on coronal radiographic images and used the slope data to estimate the Cobb angle, achieving absolute errors lower than 3°.…”

Section: Applications Of Ai and ML In Spine Researchmentioning

confidence: 99%

Artificial intelligence and machine learning in spine research

2019

View full text Add to dashboard Cite

Artificial intelligence (AI) and machine learning (ML) techniques are revolutionizing several industrial and research fields like computer vision, autonomous driving, natural language processing, and speech recognition. These novel tools are already having a major impact in radiology, diagnostics, and many other fields in which the availability of automated solution may benefit the accuracy and repeatability of the execution of critical tasks. In this narrative review, we first present a brief description of the various techniques that are being developed nowadays, with special focus on those used in spine research. Then, we describe the applications of AI and ML to problems related to the spine which have been published so far, including the localization of vertebrae and discs in radiological images, image segmentation, computer‐aided diagnosis, prediction of clinical outcomes and complications, decision support systems, content‐based image retrieval, biomechanics, and motion analysis. Finally, we briefly discuss major ethical issues related to the use of AI in healthcare, namely, accountability, risk of biased decisions as well as data privacy and security, which are nowadays being debated in the scientific community and by regulatory agencies.

show abstract

Section: Applications Of Ai and ML In Spine Researchmentioning

confidence: 99%

Artificial intelligence and machine learning in spine research

2019

View full text Add to dashboard Cite

show abstract

“…To circumvent these limitations, direct methods [9]- [15] were proposed to roughly estimate Cobb angles. Two initial attempts [8], [16] have been put forward in conference. Sun et al [16] aimed to improve the robustness of spinal curvature assessment by consolidating the tasks of vertebral landmark detection with Cobb angle estimation by exploiting the dependency between the two tasks.…”

Section: ) Direct Methodsmentioning

confidence: 99%

An Automated and Accurate Spine Curve Analysis System

Chen

Wang

et al. 2019

IEEE Access

Self Cite

View full text Add to dashboard Cite

We present a new Adaptive Error Correction Net (AEC-Net) to formulate the estimation of Cobb anges from spinal X-rays as a high-precision regression task. Our AEC-Net introduces two novel innovations. (1) The AEC-Net contains two networks calculating landmarks and Cobb angles separately, which robustly solve the disadvantage of ambiguity in X-rays since these networks focus on more features. It effectively handles the nonlinear relationship between input images and quantitative outputs, while explicitly capturing the intrinsic features of input images. (2) Based on the two estimated angles, the AEC-Net proposed a new loss function to calculate the final Cobb angles. The optimization of the loss function is based on a high-precision calculation method. The deep learning structure is used to complete this optimization, which achieves higher accuracy and efficiency. We validate our method with the spinal X-rays dataset of 581 subjects with signs of scoliosis at varying extents. The proposed method achieves high accuracy and robustness on the Cobb angle estimations. Comparing to the exsiting conventional methods suffering from tremendous variability and low reliability caused by high ambiguity and variability around boundaries of the vertebrae, the AEC-Net obtain Cobb angles accurately and robustly, which indicates its great potential in clinical use. The highly accurate Cobb angles produced by our framework can be used by clinicians for comprehensive scoliosis assessment, and possibly be further extended to other clinical applications. INDEX TERMS AEC-Net, Cobb angle estimation, deep learning, direct estimation, high-precision calculation. LIANSHENG WANG, photograph and biography not available at the time of publication.

show abstract

“…In Landmark based approach which is the state-of-the-art, the four corners of each vertebrae are detected and are subsequently used for estimating Cobb angles. Some methods jointly estimate all the landmarks and Cobb angles, while others first estimate landmarks followed by Cobb angle computation which might include outlier rejection and post-processing techniques [6,7].…”

Section: Introductionmentioning

confidence: 99%

Automatic Cobb Angle Detection Using Vertebra Detector and Vertebra Corners Regression

Khanal¹,

Dahal²,

Prabha³

et al. 2020

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Correct evaluation and treatment of Scoliosis require accurate estimation of spinal curvature. Current gold standard is to manually estimate Cobb Angles in spinal X-ray images which is time consuming and has high inter-rater variability. We propose an automatic method with a novel framework that first detects vertebrae as objects followed by a landmark detector that estimates the 4 landmark corners of each vertebra separately. Cobb Angles are calculated using the slope of each vertebra obtained from the predicted landmarks. For inference on test data, we perform pre and post processings that include cropping, outlier rejection and smoothing of the predicted landmarks. The results were assessed in AASCE MICCAI challenge 2019 which showed a promise with a SMAPE score of 25.69 on the challenge test set.

show abstract

Direct Estimation of Spinal Cobb Angles by Structured Multi-output Regression

Cited by 60 publications

References 19 publications

Artificial intelligence and machine learning in spine research

Artificial intelligence and machine learning in spine research

An Automated and Accurate Spine Curve Analysis System

Automatic Cobb Angle Detection Using Vertebra Detector and Vertebra Corners Regression

Contact Info

Product

Resources

About