Back propagation neural network(BP neural network) is a type of multi-layer feed forward network which spread positively, while the error spread backwardly. Since BP network has advantages in learning and storing the ...
详细信息
ISBN:
(数字)9781510609921
ISBN:
(纸本)9781510609914;9781510609921
Back propagation neural network(BP neural network) is a type of multi-layer feed forward network which spread positively, while the error spread backwardly. Since BP network has advantages in learning and storing the mapping between a large number of input and output layers without complex mathematical equations to describe the mapping relationship, it is most widely used. BP can iteratively compute the weight coefficients and thresholds of the network based on the training and back propagation of samples, which can minimize the error sum of squares of the network. Since the boundary of the computed tomography (CT) heart images is usually discontinuous, and it exist large changes in the volume and boundary of heart images, The conventional segmentation such as region growing and watershed algorithm can't achieve satisfactory results. Meanwhile, there are large differences between the diastolic and systolic images. The conventional methods can'tt accurately classify the two cases. In this paper, we introduced BP to handle the segmentation of heart images. We segmented a large amount of CT images artificially to obtain the samples, and the BP network was trained based on these samples. To acquire the appropriate BP network for the segmentation of heart images, we normalized the heart images, and extract the gray-level information of the heart. Then the boundary of the images was input into the network to compare the differences between the theoretical output and the actual output, and we reinput the errors into the BP network to modify the weight coefficients of layers. Through a large amount of training, the BP network tend to be stable, and the weight coefficients of layers can be determined, which means the relationship between the CT images and the boundary of heart.
A Contrast and Attenuation-map (CT-number) Linearity Improvement (CALI) framework is proposed for cone-beam CT (CBCT) images used for brain stereotactic radiosurgery (SRS). The proposed framework is used together with...
详细信息
ISBN:
(数字)9781510607101
ISBN:
(纸本)9781510607095;9781510607101
A Contrast and Attenuation-map (CT-number) Linearity Improvement (CALI) framework is proposed for cone-beam CT (CBCT) images used for brain stereotactic radiosurgery (SRS). The proposed framework is used together with our high spatial resolution iterative reconstruction algorithm and is tailored for the Leksell Gamma Knife ICON (Elekta, Stockholm, Sweden). The incorporated CBCT system in ICON facilitates frameless SRS planning and treatment delivery. The ICON employs a half-cone geometry to accommodate the existing treatment couch. This geometry increases the amount of artifacts and together with other physical imperfections causes image inhomogeneity and contrast reduction. Our proposed framework includes a preprocessing step, involving a shading and beam-hardening artifact correction, and a post-processing step to correct the dome/capping artifact caused by the spatial variations in x-ray energy generated by bowtie-filter. Our shading correction algorithm relies solely on the acquired projection images (i.e. no prior information required) and utilizes filtered-back-projection (FBP) reconstructed images to generate a segmented bone and soft-tissue map. Ideal projections are estimated from the segmented images and a smoothed version of the difference between the ideal and measured projections is used in correction. The proposed beam-hardening and dome artifact corrections are segmentation free. The CALI was tested on CatPhan, as well as patient images acquired on the ICON system. The resulting clinical brain images show substantial improvements in soft contrast visibility, revealing structures such as ventricles and lesions which were otherwise un-detectable in FBP-reconstructed images. The linearity of the reconstructed attenuation-map was also improved, resulting in more accurate CT#.
In this paper, we propose a novel supervoxel segmentation method designed for mediastinal lymph node by embedding Hessian-based feature extraction. Starting from a popular supervoxel segmentation method, SLIC, which c...
详细信息
Ovarian masses are categorised into different types of malignant and benign. In order to optimize patient treatment, it is necessary to carry out pre-operational characterisation of the suspect ovarian mass to determi...
详细信息
ISBN:
(数字)9781510609440
ISBN:
(纸本)9781510609433;9781510609440
Ovarian masses are categorised into different types of malignant and benign. In order to optimize patient treatment, it is necessary to carry out pre-operational characterisation of the suspect ovarian mass to determine its category. Ultrasound imaging has been widely used in differentiating malignant from benign cases due to its safe and non-intrusive nature, and can be used for determining the number of cysts in the ovary. Presently, the gynaecologist is tasked with manually counting the number of cysts shown on the ultrasound image. This paper proposes, a new approach that automatically segments the ovarian masses and cysts from a static B-mode image. Initially, the method uses a trainable segmentation procedure and a trained neural network classifier to accurately identify the position of the masses and cysts. After that, the borders of the masses can be appraised using watershed transform. The effectiveness of the proposed method has been tested by comparing the number of cysts identified by the method against the manual examination by a gynaecologist. A total of 65 ultrasound images were used for the comparison, and the results showed that the proposed solution is a viable alternative to the manual counting method for accurately determining the number of cysts in a US ovarian image.
Quantifying the accuracy of segmentation and manual delineation of organs, tissue types and tumors in medical images is a necessary measurement that suffers from multiple problems. One major shortcoming of all accurac...
详细信息
ISBN:
(纸本)9781510600195
Quantifying the accuracy of segmentation and manual delineation of organs, tissue types and tumors in medical images is a necessary measurement that suffers from multiple problems. One major shortcoming of all accuracy measures is that they neglect the anatomical significance or relevance of different zones within a given segment. Hence, existing accuracy metrics measure the overlap of a given segment with a ground-truth without any anatomical discrimination inside the segment. For instance, if we understand the rectal wall or urethral sphincter as anatomical zones, then current accuracy measures ignore their significance when they are applied to assess the quality of the prostate gland segments. In this paper, we propose an anatomy-aware measurement scheme for segmentation accuracy of medical images. The idea is to create a "master gold" based on a consensus shape containing not just the outline of the segment but also the outlines of the internal zones if existent or relevant. To apply this new approach to accuracy measurement, we introduce the anatomy-aware extensions of both Dice coefficient and Jaccard index and investigate their effect using 500 synthetic prostate ultrasound images with 20 different segments for each image. We show that through anatomy-sensitive calculation of segmentation accuracy, namely by considering relevant anatomical zones, not only the measurement of individual users can change but also the ranking of users' segmentation skills may require reordering.
This article describes a novel real-time algorithm for the purpose of extracting box-like structures from RGBD image data. In contrast to conventional approaches, the proposed algorithm includes two novel attributes: ...
详细信息
ISBN:
(纸本)9781510601086
This article describes a novel real-time algorithm for the purpose of extracting box-like structures from RGBD image data. In contrast to conventional approaches, the proposed algorithm includes two novel attributes: (1) it divides the geometric estimation procedure into subroutines having atomic incremental computational costs, and (2) it uses a generative "Block World" perceptual model that infers both concave and convex box elements from detection of primitive box substructures. The end result is an efficient geometry processing engine suitable for use in real-time embedded systems such as those on an UAVs where it is intended to be an integral component for robotic navigation and mapping applications.
The overall goal of this work is to develop a rapid, accurate and fully automated software tool to estimate patient-specific organ doses from computed tomography (CT) scans using a deterministic Boltzmann Transport Eq...
详细信息
ISBN:
(纸本)9781510600188
The overall goal of this work is to develop a rapid, accurate and fully automated software tool to estimate patient-specific organ doses from computed tomography (CT) scans using a deterministic Boltzmann Transport Equation solver and automated CT segmentation algorithms. This work quantified the accuracy of organ dose estimates obtained by an automated segmentation algorithm. The investigated algorithm uses a combination of feature-based and atlas-based methods. A multi atlas approach was also investigated. We hypothesize that the autosegmentation algorithm is sufficiently accurate to provide organ dose estimates since random errors at the organ boundaries will average out when computing the total organ dose. To test this hypothesis, twenty head-neck CT scans were expertly segmented into nine regions. A leave-one-out validation study was performed, where every case was automatically segmented with each of the remaining cases used as the expert atlas, resulting in nineteen automated segmentations for each of the twenty datasets. The segmented regions were applied to gold-standard Monte Carlo dose maps to estimate mean and peak organ doses. The results demonstrated that the fully automated segmentation algorithm estimated the mean organ dose to within 10% of the expert segmentation for regions other than the spinal canal, with median error for each organ region below 2%. In the spinal canal region, the median error was 7% across all data sets and atlases, with a maximum error of 20%. The error in peak organ dose was below 10% for all regions, with a median error below 4% for all organ regions. The multiple-case atlas reduced the variation in the dose estimates and additional improvements may be possible with more robust multi-atlas approaches. Overall, the results support potential feasibility of an automated segmentation algorithm to provide accurate organ dose estimates.
A unified model-based approach to ATR that uses 3D models to control detection, segmentation, and classification is described. Objects are modeled by rectangular boxes whose dimensions are Gaussian random variables. A...
详细信息
ISBN:
(纸本)9781628415902
A unified model-based approach to ATR that uses 3D models to control detection, segmentation, and classification is described. Objects are modeled by rectangular boxes whose dimensions are Gaussian random variables. A fast predictor estimates the size and shape of expected objects in the image, which controls detection and segmentation algorithms. Segmentation fits oriented rectangles (length x width @ pose) to object-like regions detected using a multi-level thresholding/region tracking approach. Detections are classified by comparing measured to predicted region length and width in the pose direction. The method is fast and requires only a coarse characterization of objects/object classes.
Deep Learning, refers to large set of neural network based algorithms, have emerged as promising machine learning tools in the general imaging and computer vision domains. Convolutional neural networks (CNNs), a speci...
详细信息
ISBN:
(纸本)9781510600201
Deep Learning, refers to large set of neural network based algorithms, have emerged as promising machine learning tools in the general imaging and computer vision domains. Convolutional neural networks (CNNs), a specific class of deep learning algorithms, have been extremely effective in object recognition and localization in natural images. A characteristic feature of CNNs, is the use of a locally connected multi layer topology that is inspired by the animal visual cortex (the most powerful vision system in existence). While CNNs, perform admirably in object identification and localization tasks, typically require training on extremely large datasets. Unfortunately, in medical image analysis, large datasets are either unavailable or are extremely expensive to obtain. Further, the primary tasks in medical imaging are organ identification and segmentation from 3D scans, which are different from the standard computer vision tasks of object recognition. Thus, in order to translate the advantages of deep learning to medical image analysis, there is a need to develop deep network topologies and training methodologies, that are geared towards medical imaging related tasks and can work in a setting where dataset sizes are relatively small. In this paper, we present a technique for stacked supervised training of deep feed forward neural networks for segmenting organs from medical scans. Each 'neural network layer' in the stack is trained to identify a sub region of the original image, that contains the organ of interest. By layering several such stacks together a very deep neural network is constructed. Such a network can be used to identify extremely small regions of interest in extremely large images, inspite of a lack of clear contrast in the signal or easily identifiable shape characteristics. What is even more intriguing is that the network stack achieves accurate segmentation even when it is trained on a single image with manually labelled ground truth. We validate th
The diaphragm is a sheet of muscle which separates the thorax from the abdomen and it acts as the most important muscle of the respiratory system. As such, an accurate segmentation of the diaphragm, not only provides ...
详细信息
ISBN:
(纸本)9781628415070
The diaphragm is a sheet of muscle which separates the thorax from the abdomen and it acts as the most important muscle of the respiratory system. As such, an accurate segmentation of the diaphragm, not only provides key information for functional analysis of the respiratory system, but also can be used for locating other abdominal organs such as the liver. However, diaphragm segmentation is extremely challenging in non-contrast CT images due to the diaphragm's similar appearance to other abdominal organs. In this paper, we present a fully automatic algorithm for diaphragm segmentation in non-contrast CT images. The method is mainly based on a priori knowledge about the human diaphragm anatomy. The diaphragm domes are in contact with the lungs and the heart while its circumference runs along the lumbar vertebrae of the spine as well as the inferior border of the ribs and sternum. As such, the diaphragm can be delineated by segmentation of these organs followed by connecting relevant parts of their outline properly. More specifically, the bottom surface of the lungs and heart, the spine borders and the ribs are delineated, leading to a set of scattered points which represent the diaphragm's geometry. Next, a B-spline filter is used to find the smoothest surface which pass through these points. This algorithm was tested on a non-contrast CT image of a lung cancer patient. The results indicate that there is an average Hausdorff distance of 2.96 mm between the automatic and manually segmented diaphragms which implies a favourable accuracy.
暂无评论