Fully Automatic X-Ray Image Segmentation via Joint Estimation of Image Displacements

Poster 139: Utilization of 3D Statistical Shape Modeling of Anatomic Variation Across a Cohort of Patella Instability Patients

Horizontal and vertical search artificial bee colony for image segmentation of COVID-19 X-ray images

We propose a method for epicardial segmentation in dynamic MR sequences using priors on shape, intensity, and dynamics. The prior models are built from a training set of manually segmented sequences. The dynamics of the heart and background intensities are modeled with linear autoregressive models whose parameters are learnt from training data and then used as priors for segmentation purposes. The segmented training shapes are registered with respect to a common reference. Each registered shape is represented with a signed distance function, and a statistical shape model is learnt using PCA on these functions. Segmentation is achieved by minimizing a spatial-temporal generalization of the Mumford-Shah energy functional in a level set framework

Pelvic X-ray (PXR) is widely utilized in clinical decision-making associated with the pelvis, the lower part of the trunk that supports and balances the trunk. In particular, PXR-based landmark detection facilitates downstream analysis and computer-assisted diagnosis and treatment of pelvic diseases. Although PXR has the advantages of low radiation and reduced cost compared to computed tomography (CT), it characterizes the 2D pelvis-tissue superposition of 3D structures, which may affect the accuracy of landmark detection in some cases. However, the superposition nature of PXR is implicitly handled by existing deep learning-based landmark detection methods, which mainly design the deep network structures for better detection performances. Explicit handling of the superposition nature of PXR is rarely done. In this paper, we explicitly focus on the superposition of X-ray images. Specifically, we propose a pelvis extraction (PELE) module that consists of a decomposition network, a domain adaptation network, and an enhancement module, which utilizes 3D prior anatomical knowledge in CT to guide and well isolate the pelvis from PXR, thereby eliminating the influence of soft tissue for landmark detection. The extracted pelvis image, after enhancement, is then used for landmark detection. We conduct an extensive evaluation based on two public and one private dataset, totaling 850 PXRs. The experimental results show that the proposed PELE module significantly improves the accuracy of PXRs landmark detection and achieves state-of-the-art performances in several benchmark metrics. The design of PELE module can improve the accuracy of different pelvic landmark detection baselines, which we believe is obviously conducive to the positioning and inspection of clinical landmarks and critical structures, thus better serving downstream tasks. Our project has been open-sourced at https://github.com/ECNUACRush/PELEscores .

The context of this work is related to the vertebra segmentation. The method we propose is based on the active shape model (ASM). An original approach taking advantage of the edge polygonal approximation was developed to locate the vertebra positions in a X-ray image. Despite the fact that segmentation results show good efficiency, the time is a key variable that has always to be optimized in a medical context. Therefore, we present how vertebra extraction can efficiently be performed in exploiting the full computing power of parallel(GPU) and heterogeneous (multi-CPU/multi-GPU) architectures. We propose a parallel hybrid implementation of the mostintensive steps enabling to boost performance. Experimentationshave been conducted using a set of high-resolution X-ray medicalimages, showing a global speedup ranging from 3 to 22, bycomparison with the CPU implementation. Data transfer timesbetween CPU and GPU memories were included in the executiontimes of our proposed implementation.

To develop a patient-specific 3 D reconstruction of a femur modeled using the statistical shape model (SSM) and X-ray images, it is assumed that the target shape is not outside the range of variations allowed by the SSM built from a training dataset. We propose the shape-partitioned statistical shape model (SPSSM) to cover significant variations in the target shape. This model can divide a shape into several segments of anatomical interest. We break up the eigenvector matrix into the corresponding representative matrices for the SPSSM by preserving the relevant rows of the original matrix without segmenting the shape and building an independent SSM for each segment. To quantify the reconstruction error of the proposed method, we generated two groups of deformation models of the femur which cannot be easily represented by the conventional SSM. One group of femurs had an anteversion angle deformation, and the other group of femurs had two different scales of the femoral head. Each experiment was performed using the leave-one-out method for twelve femurs. When the femoral head was rotated by 30°, the average reconstruction error of the conventional SSM was 5.34 mm, which was reduced to 3.82 mm for the proposed SPSSM. When the femoral head size was decreased by 20%, the average reconstruction error of the SSM was 4.70 mm, which was reduced to 3.56 mm for the SPSSM. When the femoral head size was increased by 20%, the average reconstruction error of the SSM was 4.28 mm, which was reduced to 3.10 mm for the SPSSM. The experimental results for the two groups of deformation models showed that the proposed SPSSM outperformed the conventional SSM.

Revision joint replacements are challenging surgical tasks. Knowing the exact type of primary prosthesis is essential to avoid long preoperative organisation, long operation times, and especially loss of bone and soft-tissue during operation. In daily routine there is often no information about the primary prosthesis. We are developing methods for identifying implanted prostheses from x-ray images by means of matching template images generated from prosthesis CAD data. The application is separated into three major components: The "Template Image Generation" adds 3d models of endoprostheses to a database. The "X-ray Image Segmentation" extracts endoprostheses from provided sets of x-ray images. The "Template Matching" finds the best matching prosthesis types in the data base. At the current stage, one prosthesis model (Corin, Knee ProthesisUniglide) was used for evaluating these algorithms. Very accurate identifications with accuracies of about 90% for lateral and over 70% for frontal images could be achieved. The current results of this feasibility study are very promising. A reliable and fast prosthesis identification process seems realistic to support the surgeon when planning and performing revision arthroplasty. Further improvements of segmentation accuracies and extending the prosthesis data base are intended next steps towards this goal.

We present an automatic segmentation and statistical shape modeling system for the paranasal sinuses which allows us to locate structures in and around the sinuses, as well as to observe the natural variations that occur in these structures. This system involves deformably registering a given patient image to a manually segmented template image, and using the resulting deformation field to transfer labels from template to patient. We use 3D snake splines to correct errors in the deformable registration. Once we have several accurately segmented images, we build statistical shape models for each structure in the sinus allowing us to observe the mean shape of the population, as well as the variations observed in the population. These shape models are useful in several ways. First, regular video-CT registration methods are insufficient to accurately register pre-operative computed tomography (CT) images with intra-operative endoscopy video because of deformations that occur in structures containing high amounts of erectile tissue. Our aim is to estimate these deformations using our shape models in order to improve video-CT registration, as well as to distinguish normal variations in anatomy from abnormal variations, and automatically detect and stage pathology. We can also compare the mean shape and variances of different populations, such as different genders or ethnicities, and observe the differences and similarities, as well as of different age groups, and observe the developmental changes that occur in the sinuses.

Interactive defect segmentation in X-Ray images based on deep learning

Segmentation of the left ventricular endocardium from magnetic resonance images by using different statistical shape models

This paper presents a novel variational formulation incorporating statistical knowledge to detect shapes in images. We propose to train an energy based on joint shape and feature statistics inferred from training data. Variational approaches to shape detection traditionally involve energies consisting of a feature term and a regularization term. The feature term forces the detected object to be optimal with respect to image properties such as contrast, pattern or edges whereas the regularization term stabilizes the shape of the object. Our trained energy does not rely on these two separate terms, hence avoids the non-trivial task of balancing them properly. This enables us to incorporate more complex image features while still relying on a moderate number of training samples. Cell detection in microscope images illustrates the capability of the proposed method to automatically adapt itself to different image features. We also introduce a nonlinear energy and exemplarily compare it to the linear approach.

Purpose: To build a framework for deformable B‐spline image registration with user controlled accuracy by using the positions of point landmarks as guidance. Methods: Manually identified point landmarks are placed on the reference image and their corresponding locations on the test image. An intensity‐based similarity measure is optimized under the constraint of minimizing the distance between the corresponding landmarks. After the Result of registration is visually inspected, the positions of the landmarks may be manually adjusted to improve matching accuracy. Each landmark adjustment operation triggers a re‐optimization routine, which can be repeated until the desired accuracy is reached. Results: We have tested the method on medical use cases to register CT images of the upper body, common in radiotherapy of head and neck cancer. The registration of these images compensates for position error, including pitch and yaw of the head, and flexion of the spine. Deformable registration of these images is very challenging when there is a large spatial shift of the head position. Placing landmarks on a single cross‐section and running optimization one or two times leads to a satisfactory Result. Conclusion: We have developed a practical method of interactive, intuitive correction of deformable registration of medical images. The method integrates both intensity and landmark matching. Our results prove the feasibility of the method and demonstrate potential of the tool for clinical applications.

X-ray imaging has been broadly adopted as a nondestructive testing method for product quality inspection. Deep learning has demonstrated powerful image scene understanding capabilities. In this article, U-Net with resnet101 is taken as the baseline for defect segmentation. First, there exist gray inhomogeneous and low-contrast regions in X-ray images, which can hardly be segmented. Contrast-limited adaptive histogram equalization (CLAHE) could be used to improve the contrast and consistency of the X-ray image. A two-stream convolutional neural network (CNN) is proposed that takes the original image and CLAHE processed image as inputs to address this issue. And then in CNN, low-level feature maps are lacking semantic information, which may lead to worse results. A gated multilayer fusion module is proposed to adaptively fuse the high-level features into low-level features. Furthermore, loss functions (such as cross entropy) in semantic segmentation are usually pixel level, ignoring the regional information. A weighted intersection over union (IOU) loss function is proposed to introduce IOU information to guide the model to focus on the objects that are easy to mine. The experimental results prove that the three proposed methods have better performance than the baseline for our dataset, achieving 42.2 in mIoU, 59.2 in Dice, and 54.5%, 74.9%, and 86.3% in small, middle, and large object recall rate, respectively.

Creating wax-ups of missing teeth for backward planning in implant surgery is a complex and time-consuming process. To facilitate implant-planning procedures, the automatic generation of a virtual wax-up would be useful. In the present study, the reconstruction of missing teeth in partially edentulous patients was performed automatically using newly developed software. The accuracy was investigated in order to test its clinical applicability. This study presents a new method for creating an automatic virtual wax-up, which could serve as a basic tool in modern implant-planning procedures. First, a statistical shape model (SSM) based on 76 maxillary and mandibular arch scans from dentally healthy individuals was generated. Then, artificially generated tooth gaps were reconstructed. The accuracy of the workflow was evaluated on a separate testing sample of 10 individuals with artificially created tooth gaps given as a median deviation, in millimeters. Scans of three clinical cases with partial edentulism were equally reconstructed using the SSM and compared with the final prosthodontic work. The reconstruction of the artificial tooth gaps could be performed with the following median reconstruction accuracy: gap 21 with 0.15 mm; gap 27 with 0.20 mm; gap 34 with 0.22 mm: gap 36 with 0.22 mm; gaps 12 to 22 with 0.22 mm; gaps 34 to 36 with 0.22 mm. A scenario for an almost edentulous mandible with all teeth missing except teeth 33 and 43 could be reconstructed with a median reconstruction accuracy of 0.37 mm. The median tooth gap deviation of the SSM-based reconstruction in clinical cases differed from the final inserted prosthodontic teeth by 0.49 to 0.86 mm in median. A first feasibility of creating virtual wax-ups using an SSM could be shown. Artificially generated tooth gaps could be reconstructed close to the original with the proposed workflow. In the clinical cases, the SSM proposes an anatomical reconstruction, which does not yet consider prosthodontic aspects. To obtain clinical use, contact with antagonist teeth must be considered and more training data must be implemented. However, the presented method offers a fast and viable way for the approximate placement of missing crowns. This could be used in a digital planning workflow when implant position must be determined. (Int J Comput Dent 2022;25(4):349-0; doi: 10.3290/j.ijcd.b2599407).

The construction of statistical shape models (SSMs) is an important method in the field of medical image segmentation. Most SSMs are constructed by using traditional modeling methods based on principal component analysis (PCA), which cannot fully present the true deformation ability of models. To solve the insufficient deformation ability of SSMs, we propose a stacked autoencoder (SAE) neural network to construct a multi-resolution multi-organ shape model based on mouse micro-CT images, which can express more linear and non-linear deformations than SSMs based on PCA. The main advantage of this method is that the SAE neural network is simple and flexible and it can learn more deformation modes from training data. We have quantitatively compared the modeling performance of this method with the constructed SSMs based on PCA in terms of model generalization and specificity.