Dissertations / Theses on the topic 'Image segmentation tools'

Maternal diet can have a great impact on the health and development of the fetus. Poor fetal nutrition has been linked to the development of a set of conditions in later life, such as coronary heart disease, type 2 diabetes and hypertension, while restricted growth can result in hypogylcemia, hypocalcemia, hypothermia, polycythemia, hyperbilirubinemia and cerebral palsy. High alcohol consumption during pregnancy can result in Fetal Alcohol Syndrome, a condition that can cause growth retardation, lowered intelligence and craniofacial defects. Current biometric assessment of the fetus involves size-based measures which may not accurately portray the state of fetal development, since they cannot differentiate cases of small-but-healthy or large-but-unhealthy fetuses. This thesis aims to outline a set of more appropriate measures of accurately capturing the state of fetal development. Specifically, soft tissue area and liver volume measurement are examined, followed by facial shape characterisation. A number of tools are presented which aim to allow clinicians to achieve accurate segmentations of these landmark regions. These are modifications on the Live Wire algorithm, an interactive segmentation method in which the user places a number of anchor points and a minimum cost path is calculated between the previous anchor point and the cursor. This focuses on giving the clinician intuitive control over the exact position of the segmented contour. These modifications are FA-S Live Wire, which utilises Feature Asymmetry and a weak shape constraint, ASP Live Wire, which is a 3D expansion of Live Wire, and FA-O Live Wire, which uses Feature Asymmtery and Local Orientation to guide the segmentation process. These have been designed with each of the specific biometric landmarks in mind. Finally, a method of characterising fetal face shape is proposed, using a combination of the segmentation methods described here and a simple shape model with a parameterised b-spline meshing approach to facial surface representation.

The objective of this PhD research was the development of novel 3D interactive medical platforms for medical image analysis, simulation and visualisation, with a focus on oncology images to support clinicians in managing the increasing amount of data provided by several medical image modalities. DoctorEye and Automatic Tumour Detector platforms were developed through constant interaction and feedback from expert clinicians, integrating a number of innovations in algorithms and methods, concerning image handling, segmentation, annotation, visualisation and plug-in technologies. DoctorEye is already being used in a related tumour modelling EC project (ContraCancrum) and offers several robust algorithms and tools for fast annotation, 3D visualisation and measurements to assist the clinician in better understanding the pathology of the brain area and define the treatment. It is free to use upon request and offers a user friendly environment for clinicians as it simplifies the implementation of complex algorithms and methods. It integrates a sophisticated, simple-to-use plug-in technology allowing researchers to add algorithms and methods (e.g. tumour growth and simulation algorithms for improving therapy planning) and interactively check the results. Apart from diagnostic and research purposes, it supports clinical training as it allows an expert clinician to evaluate a clinical delineation by different clinical users. The Automatic Tumour Detector focuses on abdominal images, which are more complex than those of the brain. It supports full automatic 3D detection of kidney pathology in real-time as well as 3D advanced visualisation and measurements. This is achieved through an innovative method implementing Templates. They contain rules and parameters for the Automatic Recognition Framework defined interactively by engineers based on clinicians’ 3D Golden Standard models. The Templates enable the automatic detection of kidneys and their possible abnormalities (tumours, stones and cysts). The system also supports the transmission of these Templates to another expert for a second opinion. Future versions of the proposed platforms could integrate even more sophisticated algorithms and tools and offer fully computer-aided identification of a variety of other organs and their dysfunctions.

Medical imaging technologies have allowed for in vivo evaluation of the human musculoskeletal system. With advances in both medical imaging and computing, patient-specific model development of anatomic structures is becoming a reality. Three-dimensional surface models are useful for patient-specific measurements and finite element studies. Orthopaedics is closely tied to engineering in the analysis of injury mechanisms, design of implantable medical devices, and potentially in the prediction of injury. However, a disconnection exists between medical imaging and orthopaedic analysis; whereby, the ability to generate three-dimensional models from an imaging dataset is difficult, which has restricted its application to large patient populations. We have compiled image processing, image segmentation, and surface generation tools in a single software package catered specifically to image-based orthopaedic analysis. We have also optimized an automated segmentation technique to allow for high-throughput bone segmentation and developed algorithms that help to automate the cumbersome process of mesh generation in finite element analysis. We apply these tools to evaluate graft placement in anterior cruciate ligament reconstruction in a multicenter study that aims to improve the patient outcomes of those that undergo this procedure.

Despite the association between hippocampal atrophy and a vast array of highly debilitating neurological diseases, such as Alzheimer’s disease and frontotemporal lobar degeneration, tools to accurately and robustly quantify the degeneration of this structure still largely elude us. In this thesis, we firstly evaluate previously-developed hippocampal segmentation methods (FMRIB’s Integrated Registration and Segmentation Tool (FIRST), Freesurfer (FS), and three versions of a Classifier Fusion (CF) technique) on two clinical MR datasets, to gain a better understanding of the modes of success and failure of these techniques, and to use this acquired knowledge for subsequent method improvement (e.g., FIRSTv3). Secondly, a fully automated, novel hippocampal segmentation method is developed, termed Fast Marching for Automated Segmentation of the Hippocampus (FMASH). This combined region-growing and atlas-based approach uses a 3D Sethian Fast Marching (FM) technique to propagate a hippocampal region from an automatically-defined seed point in the MR image. Region growth is dictated by both subject-specific intensity features and a probabilistic shape prior (or atlas). Following method development, FMASH is thoroughly validated on an independent clinical dataset from the Alzheimer’s Disease Neuroimaging Initiative (ADNI), with an investigation of the dependency of such atlas-based approaches on their prior information. In response to our findings, we subsequently present a novel label-warping approach to effectively account for the detrimental effects of using cross-dataset priors in atlas-based segmentation. Finally, a clinical application of MR hippocampal segmentation is presented, with a combined MR-PET analysis of wholefield and subfield hippocampal changes in Alzheimer’s disease and frontotemporal lobar degeneration. This thesis therefore contributes both novel computational tools and valuable knowledge for further neurological investigations in both the academic and the clinical field.

El processament d'imatges mèdiques és una important àrea de recerca. El desenvolupament de noves tècniques que assisteixin i millorin la interpretació visual de les imatges de manera ràpida i precisa és fonamental en entorns clínics reals.
La majoria de contribucions d'aquesta tesi són basades en Teoria de la Informació. Aquesta teoria tracta de la transmissió, l'emmagatzemament i el processament d'informació i és usada en camps tals com física, informàtica, matemàtica, estadística, biologia, gràfics per computador, etc.
En aquesta tesi, es presenten nombroses eines basades en la Teoria de la Informació que milloren els mètodes existents en l'àrea del processament d'imatges, en particular en els camps del registre i la segmentació d'imatges.
Finalment es presenten dues aplicacions especialitzades per l'assessorament mèdic que han estat desenvolupades en el marc d'aquesta tesi.
Medical imaging is an important focus of research. The development of new techniques that assist and enhance visual image interpretation in a timely and accurate manner is fundamental in real clinical environments.
Most of the contributions of this thesis are based on Information Theory. This theory deals with the transmission, storage and processing of information and it is used in fields such as physics, computer science, mathematics, statistics, economics, biology, computer graphics, etc.
In this thesis, several information theoretic tools have been presented in order to improve the existing methods in the image processing area, in particular in the image registration and segmentation fields.
Finally, two of the specialized applications for medical assessment which have been developed in the scope of this thesis are presented.

Motion segmentation refers to the process of separating regions and trajectories from a video sequence into coherent subsets of space and time. In this thesis, we created a new multifaceted motion segmentation dataset enclosing real-life long and short sequences, with different numbers of motions and frames per sequence, and real distortions with missing data. Trajectory- and region-based ground-truth is provided on all the frames of all the sequences. We also proposed a new semi-automatic tool for delineating the trajectories in complex videos, even in videos captured from moving cameras. With a minimal manual annotation of an object mask, the algorithm is able to propagate the label mask in all the frames. Object label correction based on static and moving occluder is performed by applying occluder mask tracking for a given depth ordering. The results show that our cascaded-naive approach provides successful results in a variety of video sequences.
La segmentació del moviment es refereix al procés de separar regions i trajectòries d'una seqüència de vídeo en subconjunts coherents d'espai i de temps. En aquesta tesi hem creat un nou i multifacètic dataset amb seqüències de la vida real que inclou diferent número de moviments i fotogrames per seqüència i distorsions amb dades incomplertes. A més, inclou ground-truth en tots els fotogrames basat en mesures de trajectòria i regió. Hem proposat també una nova eina semiautomàtica per delinear les trajectòries en vídeos complexos, fins i tot en vídeos capturats amb càmeres mòbils. Amb una mínima anotació manual dels objectes, l'algoritme és capaç de propagar-la en tots els fotogrames. Durant les oclusions, la correcció de les etiquetes es realitza aplicant el seguiment de la màscara per a cada ordre de profunditat. Els resultats obtinguts mostren que el nostre enfocament ofereix resultats reeixits en una àmplia varietat de seqüències de vídeo.

The human cerebellar cortex is a highly foliated structure that supports both motor and complex cognitive functions in humans. Magnetic Resonance Imaging (MRI) is commonly used to explore structural alterations in patients with psychiatric and neurological diseases. The ability to detect regional structural differences in cerebellar lobules may provide valuable insights into disease biology, progression and response to treatment, but has been hampered by the lack of appropriate tools for performing automated structural cerebellar segmentation and morphometry. In this thesis, time intensive manual tracings by an expert neuroanatomist of 16 cerebellar regions on high-resolution T1-weighted MR images of 18 children aged 9-13 years were used to generate the Cape Town Pediatric Cerebellar Atlas (CAPCA18) in the age-appropriate National Institute of Health Pediatric Database (NIHPD) asymmetric template space. An automated pipeline was developed to process the MR images and generate lobule-wise segmentations, as well as a measure of the uncertainty of the label assignments. Validation in an independent group of children with ages similar to those of the children used in the construction of the atlas, yielded spatial overlaps with manual segmentations greater than 70% in all lobules, except lobules VIIb and X. Average spatial overlap of the whole cerebellar cortex was 86%, compared to 78% using the alternative Spatially Unbiased Infra-tentorial Template (SUIT), which was developed using adult images.

Today, tool center point calibration is mostly done by a manual procedure. The method is very time consuming and the result may vary due to how skilled the operators are.

This thesis proposes a new automated iterative method for tool center point calibration of industrial robots, by making use of computer vision and image processing techniques. The new method has several advantages over the manual calibration method. Experimental verifications have shown that the proposed method is much faster, still delivering a comparable or even better accuracy. The setup of the proposed method is very easy, only one USB camera connected to a laptop computer is needed and no contact with the robot tool is necessary during the calibration procedure.

The method can be split into three different parts. Initially, the transformation between the robot wrist and the tool is determined by solving a closed loop of homogeneous transformations. Second an image segmentation procedure is described for finding point correspondences on a rotation symmetric robot tool. The image segmentation part is necessary for performing a measurement with six degrees of freedom of the camera to tool transformation. The last part of the proposed method is an iterative procedure which automates an ordinary four point tool center point calibration algorithm. The iterative procedure ensures that the accuracy of the tool center point calibration only depends on the accuracy of the camera when registering a movement between two positions.

Il concetto di identità è un concetto estremamente importante per il genere umano. L’identità, infatti, si fonda sulle caratteristiche peculiari che contraddistinguono ciascun individuo e che lo differenziano rispetto ad un altro rendendolo un soggetto unico e irripetibile. Sebbene le caratteristiche che compongano l'identità di una persona siano numerose ed estremamente differenti, si è generalmente concordi nel definire le caratteristiche fisiche come primarie nell'atto del riconoscimento personale. Per questo motivo, nella storia dell'uomo, lo sviluppo di metodologie dedicate all'identificazione si sono rivolte sempre più all'ambito fisiologico umano piuttosto che ad uno più comportamentale, culminando ai giorni nostri nei più moderni sistemi di riconoscimento biometrico. Queste tipologie di sistemi hanno assunto una dimensione pervasiva soprattutto in contesti dove i controlli umani risultato spesso complessi e limitati. Con l'avvento della pandemia di Covid-19 un numero sempre maggiore di aeroporti ed aziende ha accelerato i propri investimenti in soluzioni biometriche, motivati dalla necessità di velocizzare gli accessi minimizzando i contatti fra individui. Malgrado molti abbiano appreso questa notizia con grande entusiasmo, in letteratura esistono una serie di ricerche che mostrano come questi sistemi, sebbene siano robusti sotto scenari controllati, possano essere attualmente soggetti ad attacchi.

This thesis describes design and implementation of system that allows batch generation of graphical presentations. The system also includes modules for image quality evaluation using no-reference blur metric and salient object detection. Selected methods for evaluation of image quality are described in detail and implemented in corresponding chapters, including proposed modifications and changes. Blur detection is based on wavelet transform, and salient object detection is achieved by investigating image contrast. Capabilities of these modules are evaluated on suitable image datasets.

by Kevin Chun-Ho Wong.
Thesis (M.Phil.)--Chinese University of Hong Kong, 1998.
Includes bibliographical references (leaves 74-79).
Abstract also in Chinese.
Chinese Abstract --- p.v
Abstract --- p.vi
Acknowledgements --- p.vii
Chapter 1 --- Introduction --- p.1
Chapter 2 --- Prior Work : Image Segmentation Techniques --- p.3
Chapter 2.1 --- Introduction to Image Segmentation --- p.4
Chapter 2.2 --- Region Based Segmentation --- p.5
Chapter 2.2.1 --- Boundary Based vs Region Based --- p.5
Chapter 2.2.2 --- Region growing --- p.5
Chapter 2.2.3 --- Integrating Region Based and Edge Detection --- p.6
Chapter 2.2.4 --- Watershed Based Methods --- p.8
Chapter 2.3 --- Fuzzy Set Theory in Segmentation --- p.8
Chapter 2.3.1 --- Fuzzy Geometry Concept --- p.8
Chapter 2.3.2 --- Fuzzy C-Means (FCM) Clustering --- p.9
Chapter 2.4 --- Canny edge filter with contour following --- p.11
Chapter 2.5 --- Pyramid based Fast Curve Extraction --- p.12
Chapter 2.6 --- Curve Extraction with Multi-Resolution Fourier transformation --- p.13
Chapter 2.7 --- User interfaces for Image Segmentation --- p.13
Chapter 2.7.1 --- Intelligent Scissors --- p.14
Chapter 2.7.2 --- Magic Wands --- p.16
Chapter 3 --- Prior Work : Active Contours Model (Snakes) --- p.17
Chapter 3.1 --- Introduction to Active Contour Model --- p.18
Chapter 3.2 --- Variants and Extensions of Snakes --- p.19
Chapter 3.2.1 --- Balloons --- p.20
Chapter 3.2.2 --- Robust Dual Active Contour --- p.21
Chapter 3.2.3 --- Gradient Vector Flow Snakes --- p.22
Chapter 3.2.4 --- Energy Minimization using Dynamic Programming with pres- ence of hard constraints --- p.23
Chapter 3.3 --- Conclusions --- p.25
Chapter 4 --- Slimmed Graph --- p.26
Chapter 4.1 --- BSP-based image analysis --- p.27
Chapter 4.2 --- Split Line Selection --- p.29
Chapter 4.3 --- Split Line Selection with Summed Area Table --- p.29
Chapter 4.4 --- Neighbor blocks --- p.31
Chapter 4.5 --- Slimmed Graph Generation --- p.32
Chapter 4.6 --- Time Complexity --- p.35
Chapter 4.7 --- Results and Conclusions --- p.36
Chapter 5 --- Fast Intelligent Scissor --- p.38
Chapter 5.1 --- Background --- p.39
Chapter 5.2 --- Motivation of Fast Intelligent Scissors --- p.39
Chapter 5.3 --- Main idea of Fast Intelligent Scissors --- p.40
Chapter 5.3.1 --- Node position and Cost function --- p.41
Chapter 5.4 --- Implementation and Results --- p.42
Chapter 5.5 --- Conclusions --- p.43
Chapter 6 --- 3D Contour Detection: Volume Cutting --- p.50
Chapter 6.1 --- Interactive Volume Cutting with the intelligent scissors --- p.51
Chapter 6.2 --- Contour Selection --- p.52
Chapter 6.2.1 --- 3D Intelligent Scissors --- p.53
Chapter 6.2.2 --- Dijkstra's algorithm --- p.54
Chapter 6.3 --- 3D Volume Cutting --- p.54
Chapter 6.3.1 --- Cost function for the cutting surface --- p.55
Chapter 6.3.2 --- "Continuity function (x,y, z) " --- p.59
Chapter 6.3.3 --- Finding the cutting surface --- p.61
Chapter 6.3.4 --- Topological problems for the volume cutting --- p.61
Chapter 6.3.5 --- Assumptions for the well-conditional contour used in our algo- rithm --- p.62
Chapter 6.4 --- Implementation and Results --- p.64
Chapter 6.5 --- Conclusions --- p.64
Chapter 7 --- Conclusions --- p.71
Chapter 7.1 --- Contributions --- p.71
Chapter 7.2 --- Future Work --- p.72
Chapter 7.2.1 --- Real-time interactive tools with Slimmed Graph --- p.72
Chapter 7.2.2 --- 3D slimmed graph --- p.72
Chapter 7.2.3 --- Cartoon Film Generation System --- p.72

Video coding has been under intense scrutiny during the last years. The published international standards rely on low-level vision concepts, thus being first-generation. Recently standardization started in second-generation video coding, supported on mid-level vision concepts such as objects. This thesis presents new architectures for second-generation video codecs and some of the required analysis and coding tools. The graph theoretic foundations of image analysis are presented and algorithms for generalized shortest spanning tree problems are proposed. In this light, it is shown that basic versions of several region-oriented segmentation algorithms address the same problem. Globalization of information is studied and shown to confer different properties to these algorithms, and to transform region merging in recursive shortest spanning tree segmentation (RSST). RSST algorithms attempting to minimize global approximation error and using affine region models are shown to be very effective. A knowledge-based segmentation algorithm for mobile videotelephony is proposed. A new camera movement estimation algorithm is developed which is effective for image stabilization and scene cut detection. A camera movement compensation technique for first-generation codecs is also proposed. A systematization of partition types and representations is performed with which partition coding tools are overviewed. A fast approximate closed cubic spline algorithm is developed with applications in partition coding.
A codificação de vídeo tem sido intensamente estudada nos últimos anos. As normas internacionais já publicadas baseiam-se em conceitos da visão de baixo nível, sendo portanto de primeira geração. Começou recentemente a normalização de técnicas de codificação de segunda geração, suportada em conceitos da visão de médio nível tais como objectos. Esta tese apresenta novas arquitecturas para codificadores de vídeo de segunda geração e algumas das correspondentes ferramentas de análise e codificação. Apresentam-se fundamentos de teoria dos grafos aplicada à análise de imagem e propõem-se algoritmos para generalizações do problema da árvore abrangente mínima. Mostra-se que versões básicas de vários algoritmos de segmentação orientados para a região resolvem o mesmo problema. Estuda-se a globalização de informação e mostra-se que confere propriedades diferentes a esses algoritmos, transformando o algoritmo de fusão de regiões no algoritmo de árvores abrangentes mínimas recursivas (RSST). Mostra-se a eficácia de algoritmos RSST que tentam minimizar o erro global de aproximação e que usam modelos de região afins. Propõe-se um algoritmo baseado em conhecimento prévio para segmentação em vídeo-telefonia móvel. Desenvolve-se um algoritmo de estimação de movimentos de câmara eficaz na estabilização de imagem e na detecção de mudanças de cena. Propõe-se também uma técnica de compensação de movimentos de câmara para codificadores de primeira-geração. Sistematizam-se os tipos e as representações de regiões, revendo-se depois técnicas de codificação de partições. Desenvolve-se um algoritmo rápido e aproximado para cálculo de splines cúbicas fechadas.
Programas Ciência e Praxis - JNICT Projecto RACE MAVT - CEC ISCTE

碩士
國立臺灣大學
醫學工程學研究所
94
Object. Two aims will be achieved in the present study: 1) to develop an edge-based region growing algorithm for medical image segmentation, which has advantage of instinctive parameters adjusting; 2) to construct a framework integrated with the newly developed segmentation algorithm and an interactive image editing tool for practical application. Background. Medical image segmentation is an important image procedure for computer aided surgery and computer aided diagnosis. It is always a difficult issue to develop an algorithm for segmentation because medical images are characterized with high noise, low contrast and complex geometry. Algorithm often improves their performance in medical segmentation at the cost of increasing difficulties in internal parameters adjustment. Therefore, it is necessary to develop a user-friendly segmentation framework which also comprises the good ability to segment medical images. Method. The edge information of medical image is used as growing criterion. The boundary of grown regions was smoothed by 1D Gaussian filter at the end of each growing process. In addition to the image segmenting algorithm, interactive tools such as patching, cutting and inward region growing were developed on the basis of algorithm characteristics. To verify the efficiency of the developed algorithm and the interactive editing tools, they were applied to segment CT images and MR images. Result. The performance of algorithm was better on the regions with sufficient edge information than the regions surrounded with low contrast of soft tissue. The present developed algorithm had lower sensitivity to noise than classical region growing algorithms did. Conclusion. The region growing algorithm with good performance of medical image segmentation as well as instinctive parameters adjustment was successfully developed. Furthermore, the framework integrated with the developed algorithms and interactive editing tools can provides users with step-by-step instructions to process medical image. Therefore , users without engineering back ground can also use it easily.

Neuroanatomical data in fly brain research are mostly available as spatial gene expression patterns of genetically distinct fly strains. The Drosophila standard brain, which was developed in the past to provide a reference coordinate system, can be used to integrate these data. Working with the standard brain requires advanced image processing methods, including visualisation, segmentation and registration. The previously published VIB Protocol addressed the problem of image registration. Unfortunately, its usage was severely limited by the necessity of manually labelling a predefined set of neuropils in the brain images at hand. In this work I present novel tools to facilitate the work with the Drosophila standard brain. These tools are integrated in a well-known open-source image processing framework which can potentially serve as a common platform for image analysis in the neuroanatomical research community: ImageJ. In particular, a hardware-accelerated 3D visualisation framework was developed for ImageJ which extends its limited 3D visualisation capabilities. It is used for the development of a novel semi-automatic segmentation method, which implements automatic surface growing based on user-provided seed points. Template surfaces, incorporated with a modified variant of an active surface model, complement the segmentation. An automatic nonrigid warping algorithm is applied, based on point correspondences established through the extracted surfaces. Finally, I show how the individual steps can be fully automated, and demonstrate its application for the successful registration of fly brain images. The new tools are freely available as ImageJ plugins. I compare the results obtained by the introduced methods with the output of the VIB Protocol and conclude that our methods reduce the required effort five to ten fold. Furthermore, reproducibility and accuracy are enhanced using the proposed tools
Expressionsmuster genetisch manipulierter Fliegenstämme machen den Großteil neuroanatomischer Daten aus, wie sie in der Gehirnforschung der Taufliege Drosophila melanogaster entstehen. Das Drosophila Standardgehirn wurde u.a. entwickelt, um die Integration dieser Daten in ein einheitliches Referenz-Koordinatensystem zu ermöglichen. Die Arbeit mit dem Standardgehirn erfordert hochentwickelte Bildverarbeitungsmethoden, u.a. zur 3D Visualisierung, Segmentierung und Registrierung. Das bereits publizierte "VIB Protocol" stellte bisher eine Möglichkeit für die Registrierung zur Verfügung, die aber duch die Notwendigkeit manueller Segmentierung bestimmter Neuropile nur eingeschränkt verwendbar war. In der vorliegenden Arbeit stelle ich neue Werkzeuge vor, die den Umgang mit dem Standardgehirn erleichtern. Sie sind in ein bekanntes, offenes Bildverarbeitungsprogramm integriert, das potentiell als Standardsoftware in der neuroanatomischen Forschung dienen kann: ImageJ. Im Zuge dieser Arbeit wurde eine hardwarebeschleunigte 3D Visualisierungs-Bibliothek entwickelt, die die Visualisierungsmöglichkeiten von ImageJ ergänzt. Auf Basis dieser Entwicklung wurde anschließend ein neuer halbautomatischer Segmentierungs-Algorithmus erstellt. In diesem Algorithmus werden Neuropil-Oberflächen, ausgehend von ausgewählten Ausgangspunkten, aufgebaut und erweitert. Vorlagen von Neuropil-Oberflächen aus der Segmentierung eines Referenz-Datensatzes, die anhand eines modifizierten "Active Surface" Modells einbezogen werden können, ergänzen die aktuelle Segmentierung. Die so erhaltenen Oberflächen ermöglichen es, korrespondierende Landmarken in den Bildern zu ermitteln, die für eine nicht-rigide Registrierung verwendet werden. Schließlich wird dargelegt, wie die einzelnen Schritte voll automatisiert werden können, um die Bilder der Fliegengehirne aufeinander abzubilden. Die vorgestellten Methoden sind frei als Erweiterungen für ImageJ verfügbar (Plugins). Ein direkter Vergleich mit dem VIB Protokoll zeigt, dass durch die vorgestellten Methoden nicht nur der Benutzeraufwand auf ein Sechstel reduziert, sondern dass gleichzeitig auch die Genauigkeit und Reproduzierbarkeit erhöht wird

With the development of optical coherence tomography in the spectral domain (SD-OCT), it is now possible to quickly acquire large volumes of images. Typically analyzed by a specialist, the processing of the images is quite slow, consisting on the manual marking of features of interest in the retina, including the determination of the position and thickness of its different layers. This process is not consistent, the results are dependent on the clinician perception and do not take advantage of the technology, since the volumetric information that it currently provides is ignored. Therefore is of medical and technological interest to make a three-dimensional and automatic processing of images resulting from OCT technology. Only then we will be able to collect all the information that these images can give us and thus improve the diagnosis and early detection of eye pathologies. In addition to the 3D analysis, it is also important to develop visualization tools for the 3D data. This thesis proposes to apply 3D graphical processing methods to SD-OCT retinal images, in order to segment retinal layers. Also, to analyze the 3D retinal images and the segmentation results, a visualization interface that allows displaying images in 3D and from different perspectives is proposed. The work was based on the use of the Medical Imaging Interaction Toolkit (MITK), which includes other open-source toolkits. For this study a public database of SD-OCT retinal images will be used, containing about 360 volumetric images of healthy and pathological subjects. The software prototype allows the user to interact with the images, apply 3D filters for segmentation and noise reduction and render the volume. The detection of three surfaces of the retina is achieved through intensity-based edge detection methods with a mean error in the overall retina thickness of 3.72 0.3 pixels.

Coronal bright points (CBPs) are useful features that can be used to calculate solar rotation even when no active regions are present. Unlike active regions, CBPs are dis-tributed at all latitudes on the solar disk and its lifetime varies from less than an hour to a few days. Identifying and tracking CBPs are the main keys to successfully calculate the Solar corona rotation profile for different latitudes. Over the last years this topic has been an area of research in solar astronomy and some effective methods have been developed. The purpose of this dissertation was to design a web tool that retrieves, prepro-cesses, detects and tracks CBPs on solar images and that allows search and visualization of CBPs and solar information from a database, helping astrophysicists on their solar analysis. The detection uses a gradient based segmentation algorithm that has proved to provide accurate data about CBPs’ dynamics. It was developed a website to visualize the results, hosted by SPINLab. The track-ing from 480 images confirmed to be consistent within the expected when comparing with other authors’ work. This topic was motivated by the astrophysicists need for a near to real-time tool that allows the most recent data, as well as archive with historical data, concerning the Solar corona rotation to be processed just a few minutes after the image being captured by Nasa’s Atmospheric Imaging Assembly on board of the Solar Dynamic Observatory.

博士
義守大學
資訊工程學系
101
Chronic hepatitis is one of the most important health threats in Taiwan. Ultrasound is a widely used medical imaging technique with non-invasive safest modality. Although ultrasound is commonly used to diagnosis of various liver diseases, the accuracy of visual diagnosis is very low, and it also strongly depends on the experience of radiologist. Recently, the identification and categorization of ultrasound images have become very desirable due to the rapid development of computer technology. However, human interaction in the developed method is still inevitable. In this paper, we propose an effective technique to solve this problem. First, Otsu’s method segmentation and sampling approach are applied to sample the region of interest (ROI) from ultrasound images. Second, we propose a modification of LBP (local binary pattern) called FLBP (fuzzy local binary pattern) as texture descriptor, which is used to analyze textural features. Finally, the principle component of texture feature is introduced to classify the textures. The preliminary results indicate that the proposed methods can achieve satisfactory performance for both segmentation and categorization of liver fibrosis. In the future, the further investigation will be continued. Analysis of more ultrasound images will be performed in our experiment. Furthermore, we will also adjust the parameters to optimize the new method. We expect not only improve diagnostic accuracy of liver fibrosis but also increase the categorization accuracy.