Browsing by Author "Brijmohan, Yarish."

Now showing 1 - 2 of 2

3D modelling segmentation, quantification and visualisation of cardiovascular magnetic resonance images.
(2014) Brijmohan, Yarish.; Mneney, Stanley Henry.; Rae, William Ian Duncombe.
Progress in technology in the field of magnetic resonance imaging (MRI) has provided medical experts with a tool to visualise the heart during the cardiac cycle. The heart contains four chambers namely the left and right ventricles and the left and right atria. Each chamber plays an important role in the circulation of blood throughout the body. Imbalances in the circulatory system can lead to several cardiovascular diseases. In routine clinical medical practice MRIs are produced in large quantities on a daily basis to assist in clinical diagnosis. In practice, the interpretation of these images is generally performed visually by medical experts due to the minimal number of automatic tools and software for extracting quantitative measures. Segmentation refers to the process of detecting regions within an image and associating these regions with known objects. For cardiac MRI, segmentation of the heart distinguishes between different ventricles and atriums. If the segmentation of the left ventricle and right ventricle exists, doctors will be interested in quantifying the thickness of the ventricle walls, the movement of each ventricle, blood volumes, blood flow-rates, etc. Several cardiac MRI segmentation algorithms have been developed over the past 20 years. However, much attention of these segmentation methods was afforded to the left ventricle and its functionality due to its approximately cylindrical shape. Analysis of the right ventricle also plays an important role in heart disease assessment and coupled with left ventricle analysis, will produce a more intuitive and robust diagnostic tool. Unfortunately, the crescent like shape of the right ventricle makes its mathematical modelling difficult. Another issue associated with segmenting cardiac MRI is that the quality of images can be severely degraded by artefactual signals and image noise emanating from equipment errors, patient errors and image processing errors. The presence of these artefacts attribute to additional difficulty for segmentation algorithms and many of the currently available segmentation methods cannot account for all of the abovementioned categories. A further downfall of current segmentation algorithms is that there is no readily available standard methodology to compare the accuracy of these approaches, as each author has provided results on different cardiac MRI datasets and segmentation done by human readers (expert segmentation) is subjective. This thesis addresses the issues of accuracy comparison by providing a framework of mathematical, statistical and clinical accuracy measures. The use of publically available cardiac MRI datasets in which expert segmentation is performed is analysed. The framework allows the author of a new segmentation algorithm to choose a subset of the measures to test their algorithm. A clinical measure is proposed in this thesis which does not require expert segmentation on the cardiac MRI dataset, where the stroke volumes of the left and right ventricle are compared to each other. This thesis proposes a new three dimensional cardiac MRI segmentation algorithm that is able to segment both the left and right ventricles. This approach provides a robust technique that improves on the use of the difference of Gaussians (DoG) image filter. The main focus was to find and extract the region of interest that contains the ventricles and remove all the unwanted information so that the DoG parameters are created from intensity profiles of this localised region. Two methods are proposed to achieve this localisation, depending on the type of cardiac MRI dataset that is present. The first method is used if the cardiac MRI dataset contains images from a single MRI view. Local and global motion maps are created per MRI slice using pixel intensities from images at all time points though the cardiac cycle. The segmentation results show a slight drop in evaluation metrics on the state of the art algorithms for the left ventricle and a significant improvement over the state of the art algorithms for the right ventricle using the publically available cardiac MRI datasets. The algorithm is also robust enough to withstand the influence of image noise and simulated patient movement. The second approach to find the region of interest is used if there are MRIs from three views present in the cardiac MRI dataset. The novel method projects ventricle segmentation in the three dimensional space from two cardiac MRI views to provide an automatic ventricle localisation in the third MRI view. This method utilises an iterative approach with convergence criteria to provide final ventricle segmentation in all three MRI views. The results show increase in segmentation accuracy per iteration and a small stroke volumetric error measurement on final segmentation. Finally, proposed in this thesis is a triangular surface mesh reconstruction algorithm to create the visualisation of both the left and right ventricles. The segmentation of the ventricles are extracted from the MRI per slice and combined to form a three dimensional point set. The use of segmentation from the three orthogonal MRI views further improves the visualisation. From the three dimensional point set, the surface mesh is constructed using Delaunay triangulation, convex hulls and alpha hulls. The volume of the ventricles are calculated by performing a high resolution voxelisation of the ventricle mesh and thereafter several quantification measures are computed. The volume methodology is compared to the commonly used Simpsons method and the results illustrate that the proposed method is superior.
Codec for multimedia services using wavelets and fractals.
(2004) Brijmohan, Yarish.; Mneney, Stanley Henry.
Increase in technological advancements in fields of telecommunications, computers and television have prompted the need to exchange video, image and audio files between people. Transmission of such files finds numerous multimedia applications such as, internet multimedia, video conferencing, videophone, etc. However, the transmission and rece-ption of these files are limited by the available bandwidth as well as storage capacities of systems. Thus there is a need to develop compression systems, such that required multimedia applications can operate within these limited capacities. This dissertation presents two well established coding approaches that are used in modern' image and video compression systems. These are the wavelet and fractal methods. The wavelet based coder, which adopts the transform coding paradigm, performs the discrete wavelet transform on an image before any compression algorithms are implemented. The wavelet transform provides good energy compaction and decorrelating properties that make it suited for compression. Fractal compression systems on the other hand differ from the traditional transform coders. These algorithms are based on the theory of iterated function systems and take advantage of local self-similarities present in images. In this dissertation, we first review the theoretical foundations of both wavelet and fractal coders. Thereafter we evaluate different wavelet and fractal based compression algorithms, and assess the strengths and weakness in each case. Due to the short-comings of fractal based compression schemes, such as the tiling effect appearing in reconstructed images, a wavelet based analysis of fractal image compression is presented. This is the link that produces fractal coding in the wavelet domain, and presents a hybrid coding scheme called fractal-wavelet coders. We show that by using smooth wavelet basis in computing the wavelet transform, the tiling effect of fractal systems can be removed. The few wavelet-fractal coders that have been proposed in literature are discussed, showing advantages over the traditional fractal coders. This dissertation will present a new low-bit rate video compression system that is based on fractal coding in the wavelet domain. This coder makes use of the advantages of both the wavelet and fractal coders discussed in their review. The self-similarity property of fractal coders exploits the high spatial and temporal correlation between video frames. Thus the fractal coding step gives an approximate representation of the coded frame, while the wavelet technique adds detail to the frame. In this proposed scheme, each frame is decomposed using the pyramidal multi-resolution wavelet transform. Thereafter a motion detection operation is used in which the subtrees are partitioned into motion and non-motion subtrees. The nonmotion subtrees are easily coded by a binary decision, whereas the moving ones are coded using the combination of the wavelet SPIHT and fractal variable subtree size coding scheme. All intra-frame compression is performed using the SPIHT compression algorithm and inter-frame using the fractal-wavelet method described above. The proposed coder is then compared to current low bit-rate video coding standards such as the H.263+ and MPEG-4 coders through analysis and simulations. Results show that the proposed coder is competitive with the current standards, with a performance improvement been shown in video sequences that do not posses large global motion. Finally, a real-time implementation of the proposed algorithm is performed on a digital signal processor. This illustrates the suitability of the proposed coder being applied to numerous multimedia applications.