Department of Computer Science

Permanent link for this collection

https://hdl.handle.net/20.500.14078/7

Browse

Now showing 1 - 20 of 39

A deep level set method for image segmentation
(2017) Tang, Min; Valipour, Sepehr; Zhang, Zichen; Cobzas, Dana; Jagersand, Martin
This paper proposes a novel image segmentation approach that integrates fully convolutional networks (FCNs) with a level set model. Compared with a FCN, the integrated method can incorporate smoothing and prior information to achieve an accurate segmentation. Furthermore, different than using the level set model as a post-processing tool, we integrate it into the training phase to fine-tune the FCN. This allows the use of unlabeled data during training in a semi-supervised setting. Using two types of medical imaging data (liver CT and left ventricle MRI data), we show that the integrated method achieves good performance even when little training data is available, outperforming the FCN or the level set model alone.
A model for web-based course registration systems
(2014) Estevez, Ruben; Rankin, Sean; Silva, Ricardo; Indratmo, Indratmo
University students use web-based course registration systems to search, select, and register to courses. Despite having an important role at universities, course registration systems often pose usability problems to users. In this project, the authors assessed the usability of a web-based course registration system, proposed an improved model for such systems, and evaluated the model. The paper aims to discuss these issues.
A review of organizational structures of personal information management
(2008) Indratmo, Indratmo; Vassileva, Julita
Personal information management (PIM) covers a large area of research fragmented into separate sub-areas such as file management, web bookmark organization, and email management. Consequently, it is hard to obtain a unified view of the various approaches to PIM developed in these different sub-areas. In this article, we synthesize and classify existing research on PIM based on the approach used to organize information items. We classify the organizational structures into five categories: hierarchical, flat, linear, spatial, and network. We discuss the strengths and weaknesses of each structure along with examples showing how to deal with the weaknesses. Finally, we provide design recommendations and a framework for researchers to experiment with various ideas for developing novel PIM tools.
A usability study of an access control system for group blogs
(2007) Indratmo, Indratmo; Vassileva, Julita
Blogs are a medium to express thoughts, feelings, and opinions. Once published, blog articles potentially become persistent and can be read by non-intended audiences, causing hurt feelings and other troubles. In part these problems are due to the lack of access control in blogs. We propose an access control framework for group blogs. Compared to the typical access control in blogging tools, our system differs in a few aspects. First, the system enables bloggers to grant different access privileges to different audiences over a single blog article. That is, it associates access privileges to people rather than to artifacts (e.g., articles, blogs). Second, the system allows a blogger to create a collaborative space with other bloggers, for example by allowing others to edit his or her articles. Third, the management of access control is integrated with the process of writing and editing blog articles, facilitating the main workflow of the user. We conducted a usability study to evaluate our system and get constructive feedback from users. In this article, we present the proposed access control system, the results of the study, and analysis of the results.
An analysis of rock climbing sport regarding performance, sponsorship, and health
(2018) Huynh, Huy; Sobek, Elliott; El-Hajj, Mohamad; Atwal, Sunny
The basis of this work was to extract knowledge using data mining techniques over 2 million records from rock climbing competitions all over the world. The first phase of the project involved heavy data cleaning and preprocessing procedures to prepare the data for the mining models. After the first phase was completed, we explored three main questions: which factors can predict the performance of a competitor, which factors will likely lead to a sponsorship of a competitor and is it possible to predict healthiness of a competitor. This research will not only help rock climbers gain significant insights about their performance, it will also help sports sponsors choose the best candidates to assign their brand to.
Artificial intelligence approaches to build ticket to ride maps
(2022) Smith, Iain; Anton, Calin
Fun, as a game trait, is challenging to evaluate. Previous research explores game arc and game refinement to improve the quality of games. Fun, for some players, is having an even chance to win while executing their strategy. To explore this, we build boards for the game Ticket to Ride while optimizing for a given win rate between four AI agents. These agents execute popular strategies human players use: one-step thinking, long route exploitation, route focus, and destination hungry strategies. We create the underlying graph of a map by connecting several planar bipartite graphs. To build the map, we use a multiple phase design, with each phase implementing several simplified Monte Carlo Tree Search components. Within a phase, the components communicate with each other passively. The experiments show that the proposed approach results in improvements over randomly generated graphs and maps.
Breach path detection reliability in energy harvesting wireless sensor networks
(2021) Abougamila, Salwa; Elmorsy, Mohammed; Elmallah, Ehab S.
In this paper, we consider reliability assessment of energy harvesting wireless sensor networks (EH-WSNs) deployed to guard a geographic area against intruders that can enter and exit the network through a known set of entry-exit perimeter sides. To handle energy fluctuations during different time slots, a node may reduce its transmission power. Using a probabilistic graph model, we formalize a problem denoted EH-BPDREL (for breach path detection reliability). The problem calls for estimating the likelihood that any such intrusion can be detected and reported to a sink node. Due to the hardness of the problem, bounding algorithms are needed. We devise an efficient algorithm to solve a core problem that facilitates the design of various lower bounding algorithms. We obtain numerical results on the use of Monte Carlo simulation to estimate the probabilistic graph parameters, and illustrate the use of our devised algorithm to bound the solutions.
Comparisons between text-only and multimedia tweets on user engagement
(2020) Indratmo, Indratmo; Zhao, Michael; Buro, Karen
Having highly engaged followers on social media allows us to spread information, seek feedback, and promote a sense of community efficiently. Crafting engaging posts, however, requires careful thoughts, creativity, and communication skills. This research studied tweets and explored the effect of content types on user engagement. More specifically, we compared the number of likes and retweets between text-only and multimedia tweets. We analyzed four Twitter accounts relevant to the City of Edmonton, Canada, and performed negative binomial regressions to model the expected count of likes and retweets based on accounts, content types, and their interaction. The results showed that multimedia content increased engagement in two of the four accounts but did not change engagement significantly in the other two. In other words, multimedia content had a positive or neutral effect on user engagement, depending on accounts. Our analysis also showed the effectiveness of well-written texts in attracting the attention of users. Tweets, by design, are text-oriented, and posting multimedia content may help, but is not a necessary condition to engage with followers effectively on Twitter.
Developmental hip dysplasia diagnosis at three-dimensional US: a multicenter study
(2018) Zonoobi, Dornoosh; Hareendranathan, Abhilash; Mostofi, Emanuel; Mabee, Myles; Pasha, Saba; Cobzas, Dana; Rao, Padma; Dulai, Sukhdeep K.; Kapur, Jeevesh; Jaremko, Jacob L.
Purpose: To validate accuracy of diagnosis of developmental dysplasia of the hip (DDH) from geometric properties of acetabular shape extracted from three-dimensional (3D) ultrasonography (US).
Discriminative analysis of regional evolution of iron and myelin/calcium in deep gray matter of multiple sclerosis and healthy subjects
(2018) Elkady, Ahmed M.; Cobzas, Dana; Sun, Hongfu; Blevins, Gregg; Wilman, Alan H.
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 and demyelination. We propose and apply discriminative analysis of regional evolution (DARE) to define specific changes in MS and healthy DGM. Longitudinal (baseline and 2‐year follow‐up) retrospective study. Twenty‐seven relapsing‐remitting MS (RRMS), 17 progressive MS (PMS), and corresponding age‐matched healthy subjects . Field Strength/Sequence: 4.7T 10‐echo gradient‐echo acquisition. Automatically segmented caudate nucleus (CN), thalamus (TH), putamen (PU), globus pallidus, red nucleus (RN), substantia nigra, and dentate nucleus were retrospectively analyzed to quantify regional volumes, bulk mean R2*, and bulk mean QS. DARE utilized combined R2* and QS localized changes to compute spatial extent, mean intensity, and total changes of DGM iron and myelin/calcium over 2 years. We used mixed factorial analysis for bulk analysis, nonparametric tests for DARE (α = 0.05), and multiple regression analysis using backward elimination of DGM structures (α = 0.05, P = 0.1) to regress bulk and DARE measures with the follow‐up Multiple Sclerosis Severity Score (MSSS). False detection rate correction was applied to all tests. Bulk analysis only detected significant (Q ≤ 0.05) interaction effects in RRMS CN QS (η = 0.45; Q = 0.004) and PU volume (η = 0.38; Q = 0.034). DARE demonstrated significant group differences in all RRMS structures, and in all PMS structures except the RN. The largest RRMS effect size was CN total R2* iron decrease (r = 0.74; Q = 0.00002), and TH mean QS myelin/calcium decrease for PMS (r = 0.70; Q = 0.002). DARE iron increase using total QS demonstrated the highest correlation with MSSS (r = 0.68; Q = 0.0005). DARE enabled discriminative assessment of specific DGM changes over 2 years, where iron and myelin/calcium changes were the primary drivers in RRMS and PMS compared to age‐matched controls, respectively. Specific DARE measures of MS DGM correlated with follow‐up MSSS, and may reflect complex disease pathology.
Discriminative analysis of regional evolution of iron and myelin/calcium in deep gray matter of multiple sclerosis and healthy subjects
(2018) Elkady, Ahmed M.; Cobzas, Dana; Sun, Hongfu; Blevins, Gregg; Wilman, Alan H.
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 and demyelination. Purpose: We propose and apply discriminative analysis of regional evolution (DARE) to define specific changes in MS and healthy DGM. Study Type: Longitudinal (baseline and 2‐year follow‐up) retrospective study. Subjects: Twenty‐seven relapsing‐remitting MS (RRMS), 17 progressive MS (PMS), and corresponding age‐matched healthy subjects. Field Strength/Sequence: 4.7T 10‐echo gradient‐echo acquisition. Assessment: Automatically segmented caudate nucleus (CN), thalamus (TH), putamen (PU), globus pallidus, red nucleus (RN), substantia nigra, and dentate nucleus were retrospectively analyzed to quantify regional volumes, bulk mean R2*, and bulk mean QS. DARE utilized combined R2* and QS localized changes to compute spatial extent, mean intensity, and total changes of DGM iron and myelin/calcium over 2 years. Statistical Tests: We used mixed factorial analysis for bulk analysis, nonparametric tests for DARE (α = 0.05), and multiple regression analysis using backward elimination of DGM structures (α = 0.05, P = 0.1) to regress bulk and DARE measures with the follow‐up Multiple Sclerosis Severity Score (MSSS). False detection rate correction was applied to all tests. Results: Bulk analysis only detected significant (Q ≤ 0.05) interaction effects in RRMS CN QS (η = 0.45; Q = 0.004) and PU volume (η = 0.38; Q = 0.034). DARE demonstrated significant group differences in all RRMS structures, and in all PMS structures except the RN. The largest RRMS effect size was CN total R2* iron decrease (r = 0.74; Q = 0.00002), and TH mean QS myelin/calcium decrease for PMS (r = 0.70; Q = 0.002). DARE iron increase using total QS demonstrated the highest correlation with MSSS (r = 0.68; Q = 0.0005).Data Conclusion: DARE enabled discriminative assessment of specific DGM changes over 2 years, where iron and myelin/calcium changes were the primary drivers in RRMS and PMS compared to age‐matched controls, respectively. Specific DARE measures of MS DGM correlated with follow‐up MSSS, and may reflect complex disease pathology.
End-to-end detection-segmentation network with ROI convolution
(2018) Zhang, Zichen; Tang, Min; Cobzas, Dana; Zonoobi, Dornoosh; Jagersand, Martin; Jaremko, Jacob L.
We propose an end-to-end neural network that improves the segmentation accuracy of fully convolutional networks by incorporating a localization unit. This network performs object localization first, which is then used as a cue to guide the training of the segmentation network. We test the proposed method on a segmentation task of small objects on a clinical dataset of ultrasound images. We show that by jointly learning for detection and segmentation, the proposed network is able to improve the segmentation accuracy compared to only learning for segmentation.
Evaluation of automated computed tomography segmentation to assess body composition and mortality associations in cancer patients
(2020) Cespedes Feliciano, Elizabeth M.; Popuri, Karteek; Cobzas, Dana; Baracos, Vickie E.; Beg, Mirza Faisal; Khan, Arafat Dad; Ma, Cydney; Chow, Vincent; Chow, Vincent; Prado, Carla M.; Xiao, Jingjie; Liu, Vincent; Chen, Wendy Y.; Meyerhardt, Jeffrey; Albers, Kathleen B.; Caan, Bette J.
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 Automatic Body composition Analyser using Computed tomography image Segmentation (ABACS) software automatically segments muscle and adipose tissues to speed analysis. Here, we externally evaluate ABACS in an independent dataset. Methods Among patients with non‐metastatic colorectal (n = 3102) and breast (n = 2888) cancer diagnosed from 2005 to 2013 at Kaiser Permanente, expert raters annotated tissue areas at the third lumbar vertebra (L3). To compare ABACS segmentation results to manual analysis, we quantified the proportion of pixel‐level image overlap using Jaccard scores and agreement between methods using intra‐class correlation coefficients for continuous tissue areas. We examined performance overall and among subgroups defined by patient and imaging characteristics. To compare the strength of the mortality associations obtained from ABACS's segmentations to manual analysis, we computed Cox proportional hazards ratios (HRs) and 95% confidence intervals (95% CI) by tertile of tissue area. Results Mean ± SD age was 63 ± 11 years for colorectal cancer patients and 56 ± 12 for breast cancer patients. There was strong agreement between manual and automatic segmentations overall and within subgroups of age, sex, body mass index, and cancer stage: average Jaccard scores and intra‐class correlation coefficients exceeded 90% for all tissues. ABACS underestimated muscle and visceral and subcutaneous adipose tissue areas by 1–2% versus manual analysis: mean differences were small at −2.35, −1.97 and −2.38 cm2, respectively. ABACS's performance was lowest for the <2% of patients who were underweight or had anatomic abnormalities. ABACS and manual analysis produced similar associations with mortality; comparing the lowest to highest tertile of skeletal muscle from ABACS versus manual analysis, the HRs were 1.23 (95% CI: 1.00–1.52) versus 1.38 (95% CI: 1.11–1.70) for colorectal cancer patients and 1.30 (95% CI: 1.01–1.66) versus 1.29 (95% CI: 1.00–1.65) for breast cancer patients. Conclusions In the first study to externally evaluate a commercially available software to assess body composition, automated segmentation of muscle and adipose tissues using ABACS was similar to manual analysis and associated with mortality after non‐metastatic cancer. Automated methods will accelerate body composition research and, eventually, facilitate integration of body composition measures into clinical care.
Five year iron changes in relapsing-remitting multiple sclerosis deep gray matter compared to healthy controls
(2019) Elkady, Ahmed M.; Cobzas, Dana; Sun, Hongfu; Seres, Peter; Blevins, Gregg; Wilman, Alan H.
Relapsing-Remitting MS (RRMS) Deep Grey Matter (DGM) 5 year changes were examined using MRI measures of volume, transverse relaxation rate (R2*) and quantitative magnetic susceptibility (QS). By applying Discriminative Analysis of Regional Evolution (DARE), R2* and QS changes from iron and non-iron sources were separated. 25 RRMS and 25 age-matched control subjects were studied at baseline and 5-year follow-up. Bulk DGM mean R2* and QS of the caudate nucleus, putamen, thalamus and globus pallidus were analyzed using mixed factorial analysis (α = 0.05) with sex as a covariate, while DARE employed non-parametric analysis to study regional changes. Regression/correlation analysis was performed with disease duration and MS Severity Score (MSSS). No significant change in Extended Disability Status Score was found over 5 years (baseline = 2.4 ± 1.2; follow-up = 2.8 ± 1.3). Significant time effects were found for R2* in the caudate (Q = 0.000008; η2 = 0.36), putamen (Q = 0.0000007; η2 = 0.43), and globus pallidus (Q = 0.0000007; η2 = 0.43), while significant longitudinal effects were only found for QS in the putamen (Q = 0.002; η2 = 0.22). Significant bulk interaction was only found for thalamus volume (Q = 0.02; η2 = 0.20). Iron decrease was the only detected significant effect using DARE, and the highest significant DARE effect size was mean thalamus R2* iron decrease (Q = 0.002; η2 = 0.26). No significant correlations or regressions were demonstrated with clinical measures. Thalamic atrophy was the only bulk effect that demonstrated different rates of changes over 5 years compared to age-matched controls. DARE Iron decrease in regions of the caudate, putamen, and thalamus were prominent features in stable RRMS over 5 years.
GUI tools made easy: interact with models and explore data
(2015) Schnute, Jon; Couture-Beil, Alex; Haugh, Rowan; Kronlund, Rob; Boers, Nicholas
Provides software to facilitate the design, testing, and operation of computer models. It focuses particularly on tools that make it easy to construct and edit a customized graphical user interface (GUI). Although our simplified GUI language depends heavily on the R interface to the Tcl/Tk package, a user does not need to know Tcl/Tk. Examples illustrate models built with other R packages, including PBSmapping, PBSddesolve, and BRugs. A complete user's guide `PBSmodelling-UG.pdf' shows how to use this package effectively.
Hippocampus segmentation on high resolution dffusion MRI
(2021) Efird, Cory; Neumann, Samuel; Solar, Kevin G.; Beaulieu, Christian; Cobzas, Dana
We introduce the first hippocampus segmentation method for a novel high resolution (1×1×1mm3) diffusion tensor imaging (DTI) protocol acquired in 5.5 minutes at 3T. A new augmentation technique uses subsets of the DTI dataset to create mean diffusion weighted images (DWI) with plausible noise and contrast variations. The augmented DWI along with fractional anisotropy (FA) and mean diffusivity (MD) maps are used as inputs to a powerful convolutional neural network architecture. The method is evaluated for robustness using a second diffusion protocol.
Internationalizing the student experience through computing for social good
(2020) Aheer, Komal; Bauer, Ken; Macdonell, Cam
Information technology has connected our world and its citizens in incredible ways. Despite this connectedness, students are often isolated within the "online bubbles" of their own university, city, or country. Technology provides a great opportunity to connect them to a broader global experience. We have developed and piloted a cross-institution activity as part of an Internationalization at Home (IaH) initiative to expose first year computer science students to the concept of computing for social good in an international context. We explore how differences in culture can influence students' perceptions and approaches to computing for social good. Specifically, we had students from a Mexican and a Canadian university explore how computing for social good could be used to solve issues they faced in their communities. Students participated in surveys to propose and then rank applications for social good. The students also participated in a videoconference discussion with the students from the other school to discuss their choices. Thematic analysis revealed that the students had much more in common with each other than they had differences. Both groups not only focused on similar areas of interest, but they also tended to focus on solving issues with a local scope rather than national or global scope. Despite their cultural differences, the majority students felt they were more similar to their peers of the other culture than they were different.
Low-rank plus sparse decomposition of fMRI data with application to Alzheimer's disease
(2022) Tu, Wei; Fu, Fangfang; Kong, Linglong; Jiang, Bei; Cobzas, Dana; Huang, Chao
Studying functional brain connectivity plays an important role in understanding how human brain functions and neuropsychological diseases such as autism, attention-deficit hyperactivity disorder, and Alzheimer's disease (AD). Functional magnetic resonance imaging (fMRI) is one of the most popularly used tool to construct functional brain connectivity. However, the presence of noises and outliers in fMRI blood oxygen level dependent (BOLD) signals might lead to unreliable and unstable results in the construction of connectivity matrix. In this paper, we propose a pipeline that enables us to estimate robust and stable connectivity matrix, which increases the detectability of group differences. In particular, a low-rank plus sparse (L + S) matrix decomposition technique is adopted to decompose the original signals, where the low-rank matrix L recovers the essential common features from regions of interest, and the sparse matrix S catches the sparse individual variability and potential outliers. On the basis of decomposed signals, we construct connectivity matrix using the proposed novel concentration inequality-based sparse estimator. In order to facilitate the comparisons, we also consider correlation, partial correlation, and graphical Lasso-based methods. Hypothesis testing is then conducted to detect group differences. The proposed pipeline is applied to rs-fMRI data in Alzheimer's disease neuroimaging initiative to detect AD-related biomarkers, and we show that the proposed pipeline provides accurate yet more stable results than using the original BOLD signals.
Mapping fisheries data and spatial analysis tools
(2015) Schnute, Jon; Boers, Nicholas; Haigh, Rowan; Couture-Beil, Alex; Chabot, Denis; Grandin, Chris; Johnson, Angus; Wessel, Paul; Antonio, Franklin; Lewin-Koh, Nicholas; Bivand, Roger
This software has evolved from fisheries research conducted at the Pacific Biological Station (PBS) in `Nanaimo', British Columbia, Canada. It extends the R language to include two-dimensional plotting features similar to those commonly available in a Geographic Information System (GIS). Embedded C code speeds algorithms from computational geometry, such as finding polygons that contain specified point events or converting between longitude-latitude and Universal Transverse Mercator (UTM) coordinates. Additionally, we include `C++' code developed by Angus Johnson for the `Clipper' library. Also included are data for a global shoreline and other data sets in the public domain. The R directory `.../library/PBSmapping/doc' offers a complete user's guide, which should be consulted to use package functions effectively.
Model-based clustering and classification with the multivariate t distribution
(2016) Andrews, Jeffrey; Wickins, Jaymeson; Boers, Nicholas; McNicholas, Paul
Package ‘teigen’: Fits mixtures of multivariate t-distributions (with eigen-decomposed covariance structure) via the expectation conditional-maximization algorithm under a clustering or classification paradigm.

Browse

Browsing Department of Computer Science by Title

Results Per Page

Sort Options