A5022382297- Advanced Vision and Imaging
- Medical Image Segmentation Techniques
- Computer Graphics and Visualization Techniques
- Robotics and Sensor-Based Localization
- Advanced Neuroimaging Techniques and Applications
- Advanced Image and Video Retrieval Techniques
- Advanced Neural Network Applications
- Image and Object Detection Techniques
- 3D Shape Modeling and Analysis
- Radiomics and Machine Learning in Medical Imaging
- Medical Imaging and Analysis
- Advanced MRI Techniques and Applications
- Teleoperation and Haptic Systems
- Image Enhancement Techniques
- Multiple Sclerosis Research Studies
- Virtual Reality Applications and Impacts
- Domain Adaptation and Few-Shot Learning
- Nutrition and Health in Aging
- Optical measurement and interference techniques
- Mathematical Biology Tumor Growth
- Body Composition Measurement Techniques
- AI in cancer detection
- Advanced Image Processing Techniques
- Hip disorders and treatments
- Image Retrieval and Classification Techniques
University of Alberta
2011-2024
MacEwan University
2018-2024
Division of Undergraduate Education
2021
Athabasca University
2011
Institut national de recherche en informatique et en automatique
2005-2006
Centre Inria de l'Université Grenoble Alpes
2005-2006
The proportions of muscle and fat tissues in the human body, referred to as body composition is a vital measurement for cancer patients. Body has been recently linked patient survival onset/recurrence several types cancers numerous research studies. This paper introduces fully automatic framework segmentation from CT images estimate composition. We developed novel finite element method (FEM) deformable model that incorporates priori shape information via statistical deformation (SDM) within...
Abstract Background Body composition from computed tomography (CT) scans is associated with cancer outcomes including surgical complications, chemotoxicity, and survival. Most studies manually segment CT scans, but A utomatic B ody nalyser using C omputed image S egmentation (ABACS) software automatically segments muscle adipose tissues to speed analysis. Here, we externally evaluate ABACS in an independent dataset. Methods Among patients non‐metastatic colorectal ( n = 3102) breast 2888)...
Tumor segmentation from MRI data is an important but time consuming task performed manually by medical experts. Automating this process challenging due to the high diversity in appearance of tumor tissue, among different patients and, many cases, similarity between and normal tissue. One other challenge how make use prior information about brain. In paper we propose a variational brain algorithm that extends current approaches texture using dimensional feature set calculated registered...
Purpose To investigate subcortical gray matter segmentation using transverse relaxation rate ( R 2 *) and quantitative susceptibility mapping (QSM) apply it to voxel‐based analysis in multiple sclerosis (MS). Materials Methods Voxel‐based variation * QSM within deep was examined compared standard whole‐structure 37 MS subjects matched controls. Deep nuclei (caudate, putamen, globus pallidus, thalamus) were automatically segmented morphed onto a custom atlas based on T 1 ‐weighted images....
The ability to compute body composition in cancer patients lends itself determining the specific clinical outcomes associated with fat and lean tissue stores. For example, a wasting syndrome of advanced disease associates shortened survival. Moreover, certain compartments represent sites for drug distribution are likely determinants chemotherapy efficacy toxicity. CT images abundant, but these cannot be fully exploited unless there exist practical fast approaches quantification. Here we...
Purpose To create an automated framework for localized analysis of deep gray matter (DGM) iron accumulation and demyelination using sparse classification by combining quantitative susceptibility (QS) transverse relaxation rate (R2*) maps, evaluation DGM in multiple sclerosis (MS) phenotypes relative to healthy controls. Materials Methods R2*/QS maps were computed a 4.7T 10‐echo gradient echo acquisition from 16 clinically isolated syndrome (CIS), 41 relapsing‐remitting (RR), 40...
<h3>BACKGROUND AND PURPOSE:</h3> Deep gray matter iron accumulation is increasingly recognized in association with multiple sclerosis and can be measured vivo MR imaging. The cognitive implications of this pathology are not well-understood, especially vis-à-vis deep atrophy. Our aim was to investigate the relationships between cognition MS by using 2 imaging–based iron-susceptibility measures. <h3>MATERIALS METHODS:</h3> Forty patients (relapsing-remitting, <i>n</i> = 16; progressive, 24) 27...
We propose an attention mechanism for 3D medical image segmentation. The method, named segmentation-by-detection, is a cascade of detection module followed by segmentation module. enables region interest to come and produces set object candidates which are further used as model. Rather than dealing with the entire volume, distills information from potential region. This scheme efficient solution volumetric data it reduces influence surrounding noise. especially important low signal-to-noise...
Background Combined R2* and quantitative susceptibility (QS) has been previously used in cross-sectional multiple sclerosis (MS) studies to distinguish deep gray matter (DGM) iron accumulation demyelination. Purpose We propose apply discriminative analysis of regional evolution (DARE) define specific changes MS healthy DGM. Study Type Longitudinal (baseline 2-year follow-up) retrospective study. Subjects Twenty-seven relapsing-remitting (RRMS), 17 progressive (PMS), corresponding age-matched...
Purpose To validate accuracy of diagnosis developmental dysplasia the hip (DDH) from geometric properties acetabular shape extracted three-dimensional (3D) ultrasonography (US). Materials and Methods In this retrospective multi-institutional study, 3D US was added to conventional two-dimensional (2D) 1728 infants (mean age, 67 days; age range, 3-238 days) evaluated for DDH January 2013 December 2016. Clinical after more than 6 months follow-up normal (n = 1347), borderline (Graf IIa, later...
Tumor segmentation from MRI data is an important but time consuming task performed manually by medical experts. Automating this process challenging due to the high diversity in appearance of tumor tissue among different patients and, many cases, similarity between and normal tissue. We propose a semi-automatic interactive brain system that incorporates 2D 3D automatic tools with ability adjust operator control. The provided methods are based on energy region statistics computed available...
Developmental dysplasia of the hip (DDH) is a congenital deformity occurring in ∼3% infants. If diagnosed early most cases DDH can be effectively treated using Pavlik harness. However, current diagnosis 2D ultrasound and have high inter-operator variability. In this paper we propose method to automatically segment acetabulum bone derive geometric indices from model. proposed method, multi-scale superpixels, incorporate global local image features into Deep Learning framework obtain...
Tumor segmentation from MRI data is a particularly challenging and time consuming task. Tumors have large diversity in shape appearance with intensities overlapping the normal brain tissues. In addition, an expanding tumor can also deflect deform nearby tissue. Our work addresses these last two difficult problems. We use available modalities (T1, T1c, T2) their texture characteristics to construct multi-dimensional feature set. Further, we extract clusters which provide compact...
In this paper, we present an image-based robot incremental localization algorithm which uses a panoramic map enhanced with depth from laser range finder. The (model) contains both intensity information as well sparse 3D geometric features. By assuming motion continuity, can use the in image-model to project relevant model features, specifically vertical lines, of environment its camera coordinate frame. To determine location, first acquires image and then matches 2D features projected...
We propose a novel approach for improving level set segmentation methods by embedding the potential functions from discriminatively trained conditional random field (CRF) into energy function. The CRF terms can be efficiently estimated and lead to both discriminative local potentials edge regularizers that take account interactions among labels. Unlike discrete CRFs, use of continuous framework allows natural flexible such as shape priors. show promising experimental results method on two...