Simultaneous extraction of two adjacent bony structures in x-ray images: Application to hip joint segmentation

Ouertani, Fatma; Vázquez, Carlos; Cresson, Thierry; Guise, Jacques de

doi:10.1109/icip.2015.7351669

Cited by 7 publications

(7 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…MPA (Vincent 1998;Ouertani et al 2015) has been proven to be a robust way to detect linear features in grey images using an initial solution. Starting from an initial contour denoted C 0 , the method looks for a similar shape in a userdefined searching zone.…”

Section: Minimal Path Algorithmmentioning

confidence: 99%

“…Finally, the directed graph can be solved using dynamic programming as proposed in (Ouertani et al 2015) (Figure 4). At this stage, the MPA finds a contour which follows the gradient edges.…”

Section: D Reconstruction Algorithmmentioning

confidence: 99%

“…Indeed, radiographs are noisy and bone structures overlay at some regions. The Minimal Path Algorithm (MPA) introduced by (Vincent 1998) was successfully applied to the femoral head (Ouertani et al 2015). Also, Gaussian Process Regression (GPR) was recently proposed for shape models (Lüthi et al 2016) and can be considered as a generalisation of (Laporte et al 2003).…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Quasi-automated reconstruction of the femur from bi-planar X-rays

Girinon

Gajny

Ebrahimi

et al. 2020

Computer Methods in Biomechanics and Biomedical Engineering: Im

View full text Add to dashboard Cite

3D reconstruction from low-dose Bi-Planar X-Rays (BPXR) is a rising practice in clinical routine. However, this process is time consuming and highly depends on the user. This study aims to partially automate the process for the femur, thus decreasing reconstruction time and increasing robustness. As a training set, 50 femurs are segmented from CT scans together with 120 BPXR reconstructions. From this dataset, an initial solution for the bony contours is defined through Gaussian Process Regression (GPR), using eight digitized landmarks. This initial solution is projected on both x-rays and automatically adjusted using an adapted Minimal Path Algorithm (MPA). To evaluate this method, CT-scans were acquired from 20 cadaveric femurs. For each sample, the CTbased reconstruction is compared to the one automatically generated from the digitally reconstructed radiographs. Euclidean distances between femur reconstructions and the segmented CT data are on average 1.0 mm with a Root Mean Square Error (RMSE) of 0.8 mm. Femoral torsion errors are assessed: the bias is lower than 0.1°with a 95% confidence interval of 4.8°. The proposed method substantially improves 3D reconstructions from BPXR, as it enables a fast and reliable reconstruction, without the need for manual adjustments, which is essential in clinical routine.

show abstract

Section: Minimal Path Algorithmmentioning

confidence: 99%

Section: D Reconstruction Algorithmmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Quasi-automated reconstruction of the femur from bi-planar X-rays

Girinon

Gajny

Ebrahimi

et al. 2020

Computer Methods in Biomechanics and Biomedical Engineering: Im

View full text Add to dashboard Cite

show abstract

“…Because of the high noise, low contrast, tissue overlapping with hip bones and the narrow gap between femur and acetabulum [1], femur segmentation is a challenging task. In the previous works, segmentation error always occurs in joint regions [2].…”

Section: Introductionmentioning

confidence: 99%

“…The method applies subspace to achieve segmentation of rectangular image, but the result of segmentation is not satisfactory. Ouertani et al [1] improved hip joint segmentation by simultaneously solving a dual edge-detection problem, but the results differed between the recovered contours and the manual segmentations.…”

Section: Introductionmentioning

confidence: 99%

Segmentation of Femurs in X-ray Image with Generative Adversarial Networks

Fan¹,

Han²,

Jia³

et al. 2019

dtetr

View full text Add to dashboard Cite

Segmentation of Femur bone from X-ray images is an indispensable step in computer aided analysis of medical images and orthopaedic examinations. It is more complex than segmentation from CT and MR images, because some of the less dense related tissues are difficult to distinguish from the femur in X-ray images. We present a method based on Generative Adversarial Networks (GAN) to automatically extract the femurs from hip X-ray images. Specifically, we propose a Generative Adversarial Networks for Femur Segmentation to map hip image to femur image with nonlinear relationship and produce accurate segmentation results. In Generative Adversarial Networks for Femur Segmentation, we use the adaptive contours estimation algorithm and absolute deviations loss function to optimize neural networks. Experimental results show that our method is accurate, robust, and achieves an average dice similarity coefficient of 0.995.

show abstract