In order to analyse body composition of MRI (Magnetic Resonance Imaging) volumes, atlas-based segmentation is often used to retrieve information from specific organs or anatomical regions. The method behind this technique is to use an already segmented image volume, an atlas, to segment a target image volume by registering the volumes to each other. During this registration a deformation field will be calculated, which is applied to a segmented part of the atlas, resulting in the same anatomical segmentation in the target. The drawback with this method is that the quality of the segmentation is highly dependent on the similarity between the target and the atlas, which means that many atlases are needed to obtain good segmentation results in large sets of MRI volumes. One potential solution to overcome this problem is to create the deformation field between a target and an atlas as a sequence of small deformations between more similar bodies.  In this master thesis a new method for atlas-based segmentation has been developed, with the anticipation of obtaining good segmentation results regardless of the level of similarity between the target and the atlas. In order to do so, 4000 MRI volumes were used to create a manifold of human bodies, which represented a large variety of different body types. These MRI volumes were compared to each other and the calculated similarities were saved in matrices called similarity models. Three different similarity measures were used to create the models which resulted in three different versions of the model. In order to test the hypothesis of achieving good segmentation results when the deformation field was constructed as a sequence of small deformations, the similarity models were used to find the shortest path (the path with the least dissimilarity) between a target and an atlas in the manifold.  In order to evaluate the constructed similarity models, three MRI volumes were chosen as atlases and 100 MRI volumes were randomly picked to be used as targets. The shortest paths between these volumes were used to create the deformation fields as a sequence of small deformations. The created fields were then used to segment the anatomical regions ASAT (abdominal subcutaneous adipose tissue), LPT (left posterior thigh) and VAT (visceral adipose tissue). The segmentation performance was measured with Dice Index, where segmentations constructed at AMRA Medical AB were used as ground truth. In order to put the results in relation to another segmentation method, direct deformation fields between the targets and the atlases were also created and the segmentation results were compared to the ground truth with the Dice Index. Two different types of transformation methods, one non-parametric and one affine transformation, were used to create the deformation fields in this master thesis. The evaluation showed that good segmentation results can be achieved for the segmentation of VAT for one of the constructed similarity models. These results were obtained when a non-parametric registration method was used to create the deformation fields. In order to achieve similar results for an affine registration and to improve the segmentation of other anatomical regions, further investigations are needed.

No description available.

The study of brain nuclei in neuroimaging poses challenges owing to its small size. Many neuroimaging studies have been reported for effectively locating these nuclei and characterizing their functional connectivity with other regions of the brain. Hypothalamus, Locus Coeruleus, and Ventral Tegmental area are such nuclei found in the human brain, which are challenging to visualize owing to their size and lack of tissue contrast with surrounding regions. Resting-state functional magnetic resonance imaging (rsfMRI) analysis on these nuclei enabled researchers to characterize their connectivity with other regions of the brain. An automated method to successfully isolate voxels belonging to these nuclei is still a great challenge in the field of neuroimaging. Atlas-based segmentation is the most common method used to study the anatomy and the functional connectivity of these brain nuclei. However, atlas-based segmentation has shown inconsistency due to variation in brain atlases owing to different population studies. Therefore, in this study, we try to address the research problem of brain nuclei imaging using a clustering-based approach. Clustering-based methods separate of voxels utilizing their structural and functional homogeneity to each other. This type of method can help locate and cluster the voxels belonging to the nuclei. Elimination of erroneous voxels by the use of clustering methods would significantly improve the structural and functional analysis of the nuclei in the human brain. Since several clustering methods are available in neuroimaging studies, the goal of this study is to find a robust model that has less variability across different subjects. Non-parametrical statistical analysis was performed as functional magnetic resonance imaging (fMRI) based studies are corrupted with noise and artefact. Statistical investigation on the fMRI data helps to assess the significant experimental effects.

Quantifying tissue volumes in pediatric brains from magnetic resonance (MR) images can provide insight into etiology and onset of neurological disease. Unbiased volumetric analysis can be applied to large population studies when automated image processing is possible. Standard segmentation strategies using adult atlases fail to account for varying tissue contrasts and types associated with the rapid growth and maturational changes seen in early neurodevelopment. The goal of this project was to develop an automated pipeline and two age-specific atlases capable of providing accurate tissue classification despite these challenges. The automated pipeline consisted of a stepwise initial atlas-to-subject registration, expectation maximization (EM) atlas based segmentation, and a post-processing level set segmentation for improved white/gray matter separation. This level set segmentation is a 3D and multiphase adaptation of a 2D method intended for use on images with the types of intensity Inhomogeneities found in MR images. The initial tissue maps required to determine spatial priors for the one-year-old atlas were created by manually cleaning the results of an adult atlas and the automated pipeline. Additional tissue maps were incrementally added until the spatial priors were sufficiently representative. The neonate atlas was similarly created, starting with the one-year-old atlas.

Deep brain stimulation (DBS) surgery, a type of microelectrode-guided surgery, is an effective treatment for the movement disorders patients that can no longer be treated by medications. New rigid and non-rigid image registration methods were developed for the movement disorders patients that underwent DBS surgery. These new methods help study and analyze the brain shift during the DBS surgery and perform atlas-based segmentation of the deep brain structures for the DBS surgery planning and navigation. A diploë based rigid registration method for the intra-operative brain shift analysis during the DBS surgery was developed. The proposed method for the brain shift analysis ensures rigid registration based on diploë only, which can be treated as a rigid structure as opposed to the brain tissues. The results show that the brain shift during the DBS surgery is comparable to the size of the DBS targets and should not be neglected. This brain shift may further lengthen and complicate the DBS surgery contrary to the common belief that brain shift during the DBS surgery is not considerable. We also developed an integrated electrophysiological and anatomical atlas with eleven deep brain structures segmented by an expert, and electrophysiological data of four implant locations obtained from post-op MRI data of twenty patients that underwent DBS surgery. This atlas MR image is then non-rigidly registered with the pre-operative patient MR image, which provides initial DBS target location along with the segmented deep brain structures that can be used for guidance during the microelectrode mapping of the stereotactic procedure. The atlas based approach predicts the target automatically as opposed to the manual selection currently used. The results showed that 85% of the times, this automatic selection of the target location was closer to the target when compared to currently used technique.

Segmentation of the left and right ventricles is vital to clinical magnetic resonance imaging studies of cardiac function. A single cardiac examination results in a large amount of image data. Manual analysis by experts is time consuming and also susceptible to intra- and inter-observer variability. This leads to the urgent requirement for efficient image segmentation algorithms to automatically extract clinically relevant parameters. Present segmentation techniques typically require at least some user interaction or editing, and do not deal well with the right ventricle. This thesis presents mathematical model based methods to automatically localize and segment the left and right ventricular endocardium and epicardium in 3D cardiac magnetic resonance data without any user interaction. An efficient initialization algorithm was developed which used a novel temporal Fourier analysis to determine the size, orientation and position of the heart. Quantitative validation on a large dataset containing 330 patients showed that the initialized contours had only ~ 5 pixels (modified Hausdorff distance) error on average in the middle short-axis slices. A model-based graph cuts algorithm was investigated and achieved good results on the midventricular slices, but was not found to be robust on other slices. Instead, automated segmentation of both the left and right ventricular contours was performed using a new framework, called SMPL (Simple Multi-Property Labelled) atlas based registration. This framework was able to integrate boundary, intensity and anatomical information. A comparison of similarity measures showed the sum of squared difference was most appropriate in this context. The method improved the average contour errors of the middle short-axis slices to ~ 1 pixel. The detected contours were then used to update the 3D model using a new feature-based 3D registration method. These techniques were iteratively applied to both short-axis and long-axis slices, resulting in a 3D segmentation of the patient’s heart. This automated model-based method showed a good agreement with expert observers, giving average errors of ~ 1–4 pixels on all slices.

Segmentation of the left and right ventricles is vital to clinical magnetic resonance imaging studies of cardiac function. A single cardiac examination results in a large amount of image data. Manual analysis by experts is time consuming and also susceptible to intra- and inter-observer variability. This leads to the urgent requirement for efficient image segmentation algorithms to automatically extract clinically relevant parameters. Present segmentation techniques typically require at least some user interaction or editing, and do not deal well with the right ventricle. This thesis presents mathematical model based methods to automatically localize and segment the left and right ventricular endocardium and epicardium in 3D cardiac magnetic resonance data without any user interaction. An efficient initialization algorithm was developed which used a novel temporal Fourier analysis to determine the size, orientation and position of the heart. Quantitative validation on a large dataset containing 330 patients showed that the initialized contours had only ~ 5 pixels (modified Hausdorff distance) error on average in the middle short-axis slices. A model-based graph cuts algorithm was investigated and achieved good results on the midventricular slices, but was not found to be robust on other slices. Instead, automated segmentation of both the left and right ventricular contours was performed using a new framework, called SMPL (Simple Multi-Property Labelled) atlas based registration. This framework was able to integrate boundary, intensity and anatomical information. A comparison of similarity measures showed the sum of squared difference was most appropriate in this context. The method improved the average contour errors of the middle short-axis slices to ~ 1 pixel. The detected contours were then used to update the 3D model using a new feature-based 3D registration method. These techniques were iteratively applied to both short-axis and long-axis slices, resulting in a 3D segmentation of the patient’s heart. This automated model-based method showed a good agreement with expert observers, giving average errors of ~ 1–4 pixels on all slices.

Segmentation of the left and right ventricles is vital to clinical magnetic resonance imaging studies of cardiac function. A single cardiac examination results in a large amount of image data. Manual analysis by experts is time consuming and also susceptible to intra- and inter-observer variability. This leads to the urgent requirement for efficient image segmentation algorithms to automatically extract clinically relevant parameters. Present segmentation techniques typically require at least some user interaction or editing, and do not deal well with the right ventricle. This thesis presents mathematical model based methods to automatically localize and segment the left and right ventricular endocardium and epicardium in 3D cardiac magnetic resonance data without any user interaction. An efficient initialization algorithm was developed which used a novel temporal Fourier analysis to determine the size, orientation and position of the heart. Quantitative validation on a large dataset containing 330 patients showed that the initialized contours had only ~ 5 pixels (modified Hausdorff distance) error on average in the middle short-axis slices. A model-based graph cuts algorithm was investigated and achieved good results on the midventricular slices, but was not found to be robust on other slices. Instead, automated segmentation of both the left and right ventricular contours was performed using a new framework, called SMPL (Simple Multi-Property Labelled) atlas based registration. This framework was able to integrate boundary, intensity and anatomical information. A comparison of similarity measures showed the sum of squared difference was most appropriate in this context. The method improved the average contour errors of the middle short-axis slices to ~ 1 pixel. The detected contours were then used to update the 3D model using a new feature-based 3D registration method. These techniques were iteratively applied to both short-axis and long-axis slices, resulting in a 3D segmentation of the patient’s heart. This automated model-based method showed a good agreement with expert observers, giving average errors of ~ 1–4 pixels on all slices.

Segmentation of the left and right ventricles is vital to clinical magnetic resonance imaging studies of cardiac function. A single cardiac examination results in a large amount of image data. Manual analysis by experts is time consuming and also susceptible to intra- and inter-observer variability. This leads to the urgent requirement for efficient image segmentation algorithms to automatically extract clinically relevant parameters. Present segmentation techniques typically require at least some user interaction or editing, and do not deal well with the right ventricle. This thesis presents mathematical model based methods to automatically localize and segment the left and right ventricular endocardium and epicardium in 3D cardiac magnetic resonance data without any user interaction. An efficient initialization algorithm was developed which used a novel temporal Fourier analysis to determine the size, orientation and position of the heart. Quantitative validation on a large dataset containing 330 patients showed that the initialized contours had only ~ 5 pixels (modified Hausdorff distance) error on average in the middle short-axis slices. A model-based graph cuts algorithm was investigated and achieved good results on the midventricular slices, but was not found to be robust on other slices. Instead, automated segmentation of both the left and right ventricular contours was performed using a new framework, called SMPL (Simple Multi-Property Labelled) atlas based registration. This framework was able to integrate boundary, intensity and anatomical information. A comparison of similarity measures showed the sum of squared difference was most appropriate in this context. The method improved the average contour errors of the middle short-axis slices to ~ 1 pixel. The detected contours were then used to update the 3D model using a new feature-based 3D registration method. These techniques were iteratively applied to both short-axis and long-axis slices, resulting in a 3D segmentation of the patient’s heart. This automated model-based method showed a good agreement with expert observers, giving average errors of ~ 1–4 pixels on all slices.

京都大学 / 0048 / 新制・課程博士 / 博士(医学) / 甲第18872号 / 医博第3983号 / 新制||医||1008(附属図書館) / 31823 / 京都大学大学院医学研究科医学専攻 / (主査)教授 河野 憲二, 教授 古川 壽亮, 教授 髙橋 良輔 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DGAM