This article presents the application of the full multiresolution active shape model for the segmentation of the left ventricle in ultrasound images as well as a comparison between the classical active shape model and the full multiresolution active shape model. Our objective is to evaluate the performance of the full multiresolution framework in a complex image segmentation task such as ultrasound of the left ventricle. The accuracy of this method is evaluated through the DICE coefficient between the expert annotation and the final segmentation of the full multiresolution active shape model (FMR-ASM). The validation and training data were obtained from the CAMUS database, with 100 training and 60 validation images on which we obtained a mean DICE coefficient of 0.76 with the FMR-ASM.
Deep learning (DL) is now widely used to perform tasks involving the analysis of biomedical imaging. However, the small amounts available of annotated examples of these types of images make it difficult to use DL-based systems, since large amounts of data are required for adequate generalization and performance. For this reason, in recent years, Generative Adversarial Networks (GANs) have been used to obtain synthetic images that artificially increase the amount available. Despite this, the usual training instability in GANs, in addition to their empirical design, does not always allow for high-quality results. Through the neuroevolution of GANs it has been possible to reduce these problems, but many of these works use benchmark datasets with thousands of images, a scenario that does not reflect the real conditions of cases in which it is necessary to increase the data due to the limited amount available. In this work, cDCGAN-PSO is presented, an algorithm for the neuroevolution of GANs that adapts the concepts of the DCGAN-PSO to a conditional-DCGAN that allows the synthesis of three classes of chest X-ray images and that is trained with only 600 images of each class. The synthetic images obtained from evolved GANs show good similarity with real chest X-ray images.
Accurate automatic segmentation of the prostate in ultrasound images is still a challenging research problem. In this work, we propose the use of gray level images, constructed with a sample of gray level profiles perpendicular to the contour of the prostate. A two dimensional principal component analysis (2D PCA) was performed on a set of training contour images. The reconstruction error from the 2D PCA was used as an objective function for automatic adjustment of a point distribution model of the prostate. Our method was validated on 9 ultrasound images of the prostate and compared to the optimization of an objective function based on the mean Mahalanobis distance of a sampled gray level profile to the corresponding statistical profile model. Our new method based on a 2D PCA shows improved prostate segmentation results.
Ultrasound (US) has become one of the most common forms for medical imaging in clinical practice. It is a non-invasive and safe practice that allows obtaining images in real time. It is also a technology with important challenges such as low image quality and high variability (between manufacturers and institutions) [1]. This work aims to apply a fast and accurate deep learning architecture to detect and locate cerebellum in prenatal ultrasound images. Cerebellum biometry is used to estimate fetal age [2] and cerebellum segmentation could be applied to detect malformation [3]. YOLO (You Only Look Once) is a convolutional neural network (CNN) architecture for detection, classification and location of objects in images [4]. YOLO was innovative because it solved a regression problem to predict the location (coordinates and sizes) of bounding boxes and associated classes. We used 316 ultrasound scans of fetal brains and their respective cerebellar segmentations. From these, 78 images were randomly taken to be treated as test images and the rest were available to feed the trainings. Segmentation masks were converted to numerical descriptions of bounding boxes. To deal with small data set, transfer learning was done by initializing convolutional layers with weights pretrained on Imagenet [5]. We evaluated detection using F1 score and localization using average precision (AP) for 78 test images. Our best AP was 84.8% using 121 divisions or cells per image. Future work will focus on segmentation task assisted by localization.
The growing need to perform surgical procedures, monitoring, and intervention of greater precision have led to the development of multimodal medical imaging systems. Multimodal images are a strategy to overcome the limitations of medical imaging technologies by combining the strengths of individual modalities or technologies. In this work, we propose a low-cost multimodal system that combines 3D freehand ultrasound with fringe projection profilometry to obtain information from the external and the internal structure of an object of interest. Both modalities are referred to a single coordinate system defined in the calibration to avoid post-processing and registration of the acquired images. The freehand ultrasound calibration results are similar to those previously reported in the literature using more expensive infrared tracking systems. The calibration reproducibility at the center point of the ultrasound image was 0.6202 mm for 8 independent calibrations. We tested our system on a breast phantom with tumors. Encouraging results show the potential of the system for applications in intraoperative settings.
In this paper we propose a semi-automatic method to segment the fetal cerebellum in ultrasound images. The method is based on an active shape model which includes profiles of Hermite features. In order to fit the shape model we used a PCA of Hermite features. This model was tested on ultrasound images of the fetal brain taken from 20 pregnant women with gestational weeks varying from 18 to 24. Segmentation results compared to manual annotation show a mean Hausdorff distance of 6.85 mm using a conventional active shape model trained with gray profiles, and a mean Hausdorff distance of 5.67 mm using Hermite profiles. We conclude that the Hermite profile model is more robust in segmenting fetal cerebellum in ultrasound images.
The cerebellum is an important structure to determine the gestational age of the fetus, moreover most of the abnormalities it presents are related to growth disorders. In this work, we present the results of the segmentation of the fetal cerebellum applying statistical shape and appearance models. Both models were tested on ultrasound images of the fetal brain taken from 23 pregnant women, between 18 and 24 gestational weeks. The accuracy results obtained on 11 ultrasound images show a mean Hausdorff distance of 6.08 mm between the manual segmentation and the segmentation using active shape model, and a mean Hausdorff distance of 7.54 mm between the manual segmentation and the segmentation using active appearance model. The reported results demonstrate that the active shape model is more robust in the segmentation of the fetal cerebellum in ultrasound images.
The thickness of the nuchal fold is one of the main markers for the detection of Down syndrome during the second trimester of pregnancy. In this paper are reported our preliminary results of the automatic segmentation and measurement of the nuchal fold thickness in ultrasound images of the fetal brain. The method is based on a 2D active shape model used to segment the brain structures involved in the measurement of the nuchal fold: cerebellum; brain midline; the outer edge of the occipital plate; and the outer skin edge. The algorithm was trained and tested in 10 different ultrasound images, using leave one out cross validation. We have obtained an average difference of 0.23 mm from the expert measurement of the nuchal fold, with a standard deviation of 0.1 mm.
Three dimensional ultrasound imaging has become the main modality for fetal health diagnostics, with extensive use in fetal brain imaging. According to the fetal position and the stage of development of the fetal skull, a specific plane of image acquisition is required. In most cases for a single plane of acquisition, the image quality is limited by the shadows produced by the skull. In this work a new method for registration of multiple views of 3D ultrasound of the fetal brain is reported, which results in improved imaging of the internal brain structures. In the initial stage, texture, intensity and edge features are used, with a support vector machine (SVM) for the segmentation of the skull in each of the 3D ultrasound views to be registered. The segmentation of each skull is modelled as a set of points with the centre determined with a Gaussian mixture model, where each point is assigned a probability of membership to a Gaussian determined by the posterior probability assigned by the SVM. Our method has shown improved results compared to intensity based registration, with a 52% reduction in the target registration error (TRE), and a 39% reduction in the TRE compared to feature based registration. These are encouraging results for the future development of an automatic method for registration and fusion of multiple views of 3D fetal ultrasound.
Ultrasound (US) images of the fetal brain provide the experts with valuable indicators of the fetal development. However as the skull thickens, it obstructs the transmission of the acoustic waves, which in turn occludes the anatomy behind the thickened fetal skull. A viable option to improve the visibility of the fetal brain, before complete calcification of the skull, is the calculation of a compounded image made of different views of the same anatomical plane. In this work we report a new method for the composition of ultrasound images based on the Weighted Mean of the pixels, from different views, which correspond to each position (x, y) in the final compounded image. A support vector machine (SVM) is used to calculate the weights of each pixel from a different view, based on intensity, entropy and variance features. We present the initial test results of our method on synthetic US images of a head phantom, contaminated with speckle noise; we report the signal to noise ratio (SNR) and the normalized mutual information (NMI), for different number of views (2, 3, and 5), and compare the results against images compounded using the Mean, Root Mean Square (RMS), and Geometrical Mean composition methods. With our scheme we were able to recover the occluded information to increase the NMI from 16% to 26%, representing a 58% improvement.
The cerebellum is an important structure to determine the gestational age, cerebellar diameter obtained by ultrasound volumes of the fetal brain has shown a high correlation with gestational age, therefore is useful to determine fetal growth restrictions. The manual annotation of 3D surfaces from the fetal brain is time consuming and needs to be done by a highly trained expert. To help with the annotation in the evaluation of cerebellar diameter, we developed a new automatic scheme for the segmentation of the 3D surface of the cerebellum in ultrasound volumes, using a spherical harmonics model and the optimization of an objective function based on gray level voxel profiles. The results on 10 ultrasound volumes of the fetal brain show an accuracy in the segmentation of the cerebellum (mean Dice coefficient of 0.7544). The method reported shows potential to effectively assist the experts in the assessment of fetal growth in ultrasound volumes. We consider the proposed cerebellum segmentation method a contribution for the SPHARM segmentations models, because it is capable to run without hardware restriction, (GPU), and gives adequate results in a reasonable amount of time.
We report a new method for adjusting the points of an active shape model (ASM) to the edge of an object, on a grey level image. The method is based on the original iterative search for an optimum location of each point of the ASM, along the normal direction to the model boundary. In this work we determine the optimum location of the model boundary point through minimization of the error (euclidean distance) between a profile of pixels sampled along the normal direction, and its projection on the principal component axes, obtained from a training set of normal pixel profiles, located at the edge of the object. We validated our method on ultrasound images of the prostate and photographs of the left hand. Significant improvements were observed in the segmentation of the ultrasound images, with reference to the original ASM method of adjustment, while no significant improvement was observed for the photographs. Our method produced a mean error of 4.58 (mm) between corresponding expert and automatically annotated contours of the ultrasound images of the prostate, and 3.12 (mm) of mean error for the photographs of the left hand.
We present a discrete compactness (DC) index, together with a classification scheme, based both on the size and shape features extracted from brain volumes, to determine different aging stages: healthy controls (HC), mild cognitive impairment (MCI), and Alzheimer’s disease (AD). A set of 30 brain magnetic resonance imaging (MRI) volumes for each group was segmented and two indices were measured for several structures: three-dimensional DC and normalized volumes (NVs). The discrimination power of these indices was determined by means of the area under the curve (AUC) of the receiver operating characteristic, where the proposed compactness index showed an average AUC of 0.7 for HC versus MCI comparison, 0.9 for HC versus AD separation, and 0.75 for MCI versus AD groups. In all cases, this index outperformed the discrimination capability of the NV. Using selected features from the set of DC and NV measures, three support vector machines were optimized and validated for the pairwise separation of the three classes. Our analysis shows classification rates of up to 98.3% between HC and AD, 85% between HC and MCI, and 93.3% for MCI and AD separation. These results outperform those reported in the literature and demonstrate the viability of the proposed morphological indices to classify different aging stages.
Computer Assisted Orthopedic Surgery (CAOS) requires a correct registration between the patient in the operating room and the virtual models representing the patient in the computer. In order to increase the precision and accuracy of the registration a set of new techniques that eliminated the need to use fiducial markers have been developed. The majority of these newly developed registration systems are based on costly intraoperative imaging systems like Computed Tomography (CT scan) or Magnetic resonance imaging (MRI). An alternative to these methods is the use of an Ultrasound (US) imaging system for the implementation of a more cost efficient intraoperative registration solution. In order to develop the registration solution with the US imaging system, the bone surface is segmented in both preoperative and intraoperative images, and the registration is done using the acquire surface. In this paper, we present the a preliminary results of a new approach to segment bone surface from ultrasound volumes acquired by means 3D freehand ultrasound. The method is based on the enhancement of the voxels that belongs to surface and its posterior segmentation. The enhancement process is based on the information provided by eigenanalisis of the multiscale 3D Hessian matrix. The preliminary results shows that from the enhance volume the final bone surfaces can be extracted using a singular value thresholding.
Image-guided interventions allow the physician to have a better planning and visualization of a procedure. 3D freehand ultrasound is a non-invasive and low-cost imaging tool that can be used to assist medical procedures. This tool can be used in the diagnosis and treatment of breast cancer. There are common medical practices that involve large needles to obtain an accurate diagnosis and treatment of breast cancer. In this study we propose the use of 3D freehand ultrasound for planning and guiding such procedures as core needle biopsy and radiofrequency ablation. The proposed system will help the physician to identify the lesion area, using image-processing techniques in the 3D freehand ultrasound images, and guide the needle to this area using the information of position and orientation of the surgical tools. We think that this system can upgrade the accuracy and efficiency of these procedures.
Considering the importance of studying the movement of certain cardiac structures such as left ventricle and myocardial wall for better medical diagnosis, we propose a method for motion estimation and image segmentation in sequential Computed Tomography images. Two main tasks are tackled. The first one consists of a method to estimate the heart's motion based on a bio-inspired image representation model. Our proposal for optical flow estimation incorporates image structure information extracted from the steered Hermite transform coefficients that is later used as local motion constraints in a differential estimation approach. The second task deals with cardiac structure segmentation in time series of cardiac images based on deformable models. The goal is to extend active shape models (ASM) of 2D objects to the problem of 3D (2D + time) cardiac CT image modeling. The segmentation is achieved by constructing a point distribution model (PDM) that encodes the spatio-temporal variability of a training set. Combination of both motion estimation and image segmentation allows isolating motion in cardiac structures of medical interest such as ventricle walls.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.