PRISM :: Browsing by Author "Fear, Elise"

Browsing by Author "Fear, Elise"

Now showing 1 - 20 of 44

Open Access
A coupled eulerian-langrangian mechanical model of the breast
(2010) Kuhlmann, Martin; Ramirez-Serrano, Alejandro; Fear, Elise; Federico, Salvatore
Open Access
A cross-polarized antenna for breat tumor detection
(2006) Zhang, Jingjing; Fear, Elise; Johnston, Ronald H.
Open Access
A dielectric filled slotline bowtie antenna for breast cancer detection
(2004) Shannon, Christopher James; Okoniewski, Michal; Fear, Elise
Open Access
A flush mounted waveguide antenna for telemetry applications
(2011) Cameron, Trevor R.; Okoniewski, Michal; Fear, Elise
Open Access
A microwave interferometer to enhance sensitivity of microwave measurement systems
(2012) Kagan, Dmitri; Fear, Elise
A single source microwave interferometer is proposed to enhance the sensitivity of a microwave measurement system. Theory of operation is provided to gain understanding of the device's generated null and the concept of a virtual match. The design and steps leading to its realization are discussed and developed through a comprehensive analysis involving simulations and prototyping. The performance of the interferometer is demonstrated through multiple applications, involving low contrast and high impedance measurements. Due to the ability of the interferometer to easily integrate with multiple measurement systems, it has potential benefiting applications that require extreme sensitivity and ones relying on differential information.
Open Access
Adapting Seismic Processing Techniques for Data Preconditioning in Radar Imaging of Highly Dissipative and Dispersive Media
(2017) Liu, Yuhong; Fear, Elise; Potter, Mike; Smith, Mike; Ferguson, Robert; Lines, Laurence; Popovic, Milica
The concept of using microwave frequency electromagnetic waves for biomedical imaging applications has interested researchers for decades. Promising results have been reported for several approaches to microwave breast imaging, including radar-based imaging applied to realistic numerical breast phantoms and patient studies. However, important problems have also been identified, specifically, low image resolution and sensitivity due to multiple-scattering effects and frequency-dependent attenuation in the presence of highly dissipative and dispersive breast tissues. Microwave imaging and seismic imaging deal with analogous problems. In seismic imaging, tremendous efforts have been invested in developing data analysis and preconditioning techniques to render the accurate graphical representation of specific portions of the earth’s subsurface geological structure. The overall objective of this thesis is to produce more accurate microwave breast images from ultra-wideband radar signals by adapting advanced seismic imaging techniques. First, we develop a method based on first-breaks to detect the pulse arrival time in the presence of severe waveform distortion. Second, we adapt Gabor nonstationary deconvolution to accurately estimate the subsurface reflectivity in the presence of severe attenuation and dispersion due to EM wave propagation in highly lossy dispersive biological tissues at microwave frequencies. Third, we develop a dual deconvolution processing flow (DDPF) to account for the interfering responses present in a radar reflection measurement system. The proposed methods are applied to simulated and measured data. The results indicate that the first-break time is able to provide consistent and reliable reference for travel time estimation in the presence of severe waveform distortion and Gabor deconvolution is able to effectively compensate for wave attenuation in highly lossy and dispersive media. The preliminary imaging test demonstrated a significant improvement in the image sensitivity with Gabor deconvolution preconditioned data. Application to the simulations of realistic breast phantoms and experimental patient scans shows that the DDPF method is able to detect the scatterers in the presence of heterogeneous, lossy, and dispersive tissues. Overall, this study demonstrates successful modification of seismic data preconditioning techniques to biomedical radar data, resulting in images with improved accuracy.
Open Access
Advanced cardiac imaging analysis in atrial fibrillation and its hemodynamics
(2023-12-20) Kim, Hansuk; Garcia Flores, Julio; Fear, Elise; Wilton, Stephen B.
Atrial fibrillation (AF) is a prevalent cardiac disorder characterized by rapid and disorganized atrial activation, resulting in impaired atrial function. With a global impact affecting approximately 33.5 million individuals, AF leads to an elevated risk of complications such as thromboembolism and stroke. Current treatment guidelines primarily rely on patient demographics and medical history, yet a more precise therapeutic approach based on individual flow is required. 4D flow magnetic resonance imaging (MRI) offers comprehensive measurement of flow velocity throughout the entire cardiac cycle in three dimensions. While previous studies have explored the application of 4D flow techniques in AF, they have encountered limitations related to resolution and contrast, particularly when segmenting the fine structures of the left atrium (LA). Furthermore, the clinical parameters derived from 4D flow MRI in AF have been relatively limited, and there has been notable absence of studies focused on assessing flow recovery after catheter ablation. This research addresses the technical and clinical gaps in previous studies, aiming to enhance the clinical applicability of 4D flow imaging techniques in AF. In clinical aspect, our investigation explores the impact of AF on left ventricular (LV) flow in paroxysmal AF using LV flow component analysis. This analysis unveils subtle alterations in LV flow efficiency, even in patients with paroxysmal AF and normal systolic function, indicating hemodynamic changes without signs of adverse LV remodeling. Our comparative study of LA flow in AF patients before and after catheter ablation reveals significant changes in LA blood flow stasis using 4D flow imaging. To overcome the limitations of resolution and contrast of 4D flow, we propose a segmentation method utilizing standard-of-care contrast-enhanced magnetic resonance angiography (CE-MRA) and a registration process with 4D flow data. In conclusion, this thesis advances the application of 4D-flow MRI in AF studies, offering novel insights and clinically relevant findings. These findings pave the way for more precise and individualized therapeutic strategies in the management of AF.
Open Access
Advanced Delta-Sigma Transmitter Architectures for High Performance Wireless Applications
(2017) Jouzdani, Maryam; Ghannouchi, Fadhel M.; Helaoui, Mohamed; Belostotski, Leonid; Fear, Elise; Baudoin, Geneviève; Nowicki, Edwin
To satisfy the wireless market’s growing demand for higher data rates services and to maximize the bandwidth spectral efficiency, modern modulation schemes have been developed. Transmitting spectrally efficient non-constant envelope signals modulated by modern schemes necessitates designing highly linear and efficient transmitter systems for reaching the signal-to-noise ratio (SNDR) requirements and longer battery life. Delta-sigma modulator (DSM) based transmitters have the potential of good linearity performance and re-configurability for multi-standard applications. They also enable the use of high efficiency switching power amplifiers (PAs). This thesis was dedicated to enhancing the performance of DSM based transmitters. The first part of the thesis will focus on the design and evaluation of a novel high-pass (HP) DSM- based digital-IF transmitter architecture to address the in-band quantization noise problem and low coding efficiency in Cartesian HP and band-pass (BP) counterparts. As the most power consuming part of the transmitters, the design of highly efficient RF PAs has been the subject of several studies with different techniques being proposed to overcome this challenge. Dynamic control of the load impedance of the amplifier is a promising technique used in pulsed load modulation (PLM) PAs. Digital load modulation is realized in PLM PAs with the aid of the envelope delta-sigma modulator (EDSM) to enhance the efficiency in larger power back-off region while preserving the quality of the signal. The design and fabrication of a PLM PA with gate bias modulation for high power applications is the subject of the second part of this thesis. Employing the designed PLM PA, a digital DSM-based transmitter topology was realised for base-band applications. The transmitter was successfully tested with standard signals showing promising results. In the next step, it is shown that to further increase the efficiency of the PLM PA-based transmitters, it is possible to reduce the delta-sigma quantization noise and thus, the quality of the encoded signal by replacing the EDSM by a complex delta sigma modulator (CDSM). Based on this method, a novel transmitter architecture is proposed which benefits from CDSMs and PLM PAs for reaching the SNDR requirements and high efficiency performance at the same time.
Open Access
Algorithms for the reduction of clutter in tissue sensing adaptive radar (TSAR) signals
(2007) Kurrant, Douglas John; Westwick, David; Fear, Elise
Open Access
Artifact Reduction Strategy for Radar-based Microwave Imaging Designed for Medical Applications
(2021-08-24) DasGupta, Ishani; Smith, Mike; Fear, Elise; Curiel, Laura; Far, Behrouz
This thesis involves investigating the frequency domain data obtained from Tissue Sensing Adaptive Radar (TSAR), which is a near-field ultra-wideband radar imaging technique using microwaves that has potential as a new breast imaging modality. Domain transformations of the acquired data result in Gibbs’ distortions which can propagate through the data processing flow. These distortions or artifacts can be reduced by filtering the frequency data but with a loss in time domain resolution. Fourier Shift Manipulation (FSM) was explored as an alternative pre-processing technique that utilizes fundamental discrete Fourier transform (DFT) properties to shift the sampling locations of the signal, leading to artifact reduction with minimal loss of resolution. The extent of the removal of Gibbs’ artifacts led to investigations into specific improvements in the initial time-domain signals which were subsequently used for image formation. Further exploration involved the propagation of these distortions through the data flow and the impact they have on the clutter response in the final reconstructed images. The artifact reduction techniques were initially tested on simple simulated models, then extended to more complex datasets and finally patient data. Existing metrics were used to compare the outputs from these approaches, and new ones developed wherever necessary. The differences in the resulting images were compared with an emphasis on the degree of tumour detection. Future research would involve evaluating the compatibility of FSM with other image reconstruction algorithms, as well as modifying the removal of an inherent skin-breast artifact.
Open Access
Automated detection of differences in treated and untreated breast tissue through analysis of microwave imaging data
(2022-07) Garland, Anita; Yanushkevich, Svetlana; Fear, Elise; Bayat, Sayeh; Far, Behrouz
Although microwave imaging has been researched in various applications for about 40 years, its use in biomedical applications is a more recent endeavor. In this thesis, we examine the feasibility of using automated detection on microwave imaging data for providing treatment-related feedback on changes in breast tissue. 16 female patients at the Tom Baker Cancer Centre in Calgary, were recruited from a clinical trial, to be scanned by the University of Calgary's Microwave Imaging Transmission System (MITS) for a maximum of 4 times, over a period of 2 years. Early stage breast cancer treatment typically involves lumpectomy and 1-5 weeks of radiotherapy to the breast that contained the tumour. Our hypothesis is that changes in breast tissue due to cancer treatment may be detected through analysis of microwave imaging data using machine learning algorithms. Data analysis began with exploration of the microwave frequency properties of tissue or tissue permittivity to find differences between treated and untreated tissue. Challenges of identifying specific and consistent changes across the group of patients using permittivity analysis led to switching our approach to analysis of the underlying time-domain signals. Employing wavelet transforms on the time-domain signals resulted in more defined differences between the treated and the untreated breast for feature extraction. Next, classifiers like Support Vector Machine, Random Forest and Gradient Boosting Classifier were used on the extracted features. A final analysis of the frequency domain signals and combined time-frequency domain features was also undertaken to highlight differences and apply classification to the extracted features. This thesis provides a framework for an automated technique to detect changes between treated and untreated breast tissue using the microwave scan data. Our results indicate that this approach to analyzing microwave imaging data may have the potential to extract differences in breast tissue arising from radiotherapy and/or surgery.
Open Access
Average Dielectric Property Analysis of Non-Uniform Structures: Tissue Phantom Development, Ultra-Wideband Transmission Measurements, and Signal Processing Techniques
(2014-09-12) Garrett, John Daniel; Fear, Elise
A new technique to analyze the average dielectric properties of complex structures has been developed. This technique uses microwave transmission measurements to estimate the complex permittivity of the object-under-test over a range of frequencies. First, time-gating is used to reduce multipath. Second, the antenna response is deconvolved from the measured data. Finally, with the object-under-test's response isolated, the average properties are estimated. To test and validate this work, dielectric materials with properties representing biological tissues and robust mechanical properties were developed. Reconfigurable tissue phantoms were created from combinations of these materials. In addition to testing the average property algorithms, these phantoms are well suited to testing a variety of microwave imaging methods. Complex breast phantoms were used to test the average dielectric property estimation technique, and accurate results were found. This technique was then applied to measurements of human breast tissue, and reasonable properties were estimated.
Open Access
Biometric-Enabled Decision Support Platform with Risk Assessment
(2022-01-14) Lai, Kenneth; Yanushkevich, Svetlana; Hatzinakos, Dimitrios; Hemmati, Hadi; Fear, Elise; Nielsen, John
Biometric-based human trait and behavior recognition is a critical component of the rapidly growing domain of ambient intelligence. Particular applications of interest are biometric-enabled border checkpoints, access control, as well as healthcare and biomedical data analysis. In this thesis, we offer both theoretical and practical contributions. The main theoretical contributions include the framework for uncertainty measures and performance assessment measures in a decision support system. These measures include risk, trust, and bias and the methodology involves using these measures in a contemporary engine for decision support systems, based on causal models of uncertainty. These models allow for the prediction of events of interest and assess the risks associated with these events. Our main practical contributions include the advanced practical implementation of various machine learning approaches, mostly deep neural networks, to biometric-enabled applications such as facial recognition, action recognition, emotion classification, wearable data analysis for healthcare, and human-machine interaction applications. Demonstration of practical applications of machine reasoning for biometric-enabled systems that use facial recognition, action recognition, and emotion recognition is shown. In this research, we propose to combine multi-spectral biometric data processing, powerful deep learning techniques, along with performance improvement techniques, in a unified approach to automate face and action recognition. When combined, a platform consisting of powerful machine-learning techniques is used as a supporting tool to provide decision support for an operator. Multi-spectral data is understood as color and depth data such as video, depth, and derived skeleton joints. The deep learning techniques include convolutional and recurrent neural networks. The former is applied in our study to extract important spatial information from color and depth images, whereas the latter is utilized to recognize temporal patterns. Emerging deep learning model architectures are explored, one such network called the Residual Temporal Convolutional network offers improved performance in comparison to recurrent neural networks. This research will show the capability of using different types of data to train neural networks independently to recognize biometric patterns including actions and faces. Therefore, the main focus of this thesis is the development of a decision support platform for solving a variety of practical problems. These solutions are approached using the same methodology: advanced machine learning techniques are used to process, analyze and classify data, and machine reasoning is used as a common platform for the system-level decision making, with risk, bias, and/or trust assessment of the provided decision. This approach is embodied, in particular, in a proposed decision support system for human stress detection using physiological signals. Deep learning techniques were applied for detecting and recognizing different emotional states. The causal models were built upon the distribution of the detection and recognition scores in order to perform machine reasoning and information fusion. These results provided the operator with the risk assessment of any given scenario. Other examples of such systems are provided in multiple publications as reported in the thesis.
Open Access
Breast Tissue-Mimicking Phantom for Combined Ultrasound and Microwave Imaging
(2021-03-16) Li, Siyun; Curiel, Laura; Fear, Elise; Di Martino, Elena
Breast cancer is the most common cancer among women and persists to be one of the top threats to women’s health. Early diagnosis and treatment are the keys to defeat breast cancer. There are plenty of methods and devices, such as X-ray, CT, and MRI, to detect tumours and malignant breast tissues. More recently, ultrasound imaging has become a standard cancer detection method, and microwave imaging has gained interest in diagnosing breast tumours for its moderate biological effects. Using a combination of these two modalities can provide high-resolution images with improved contrast that can make up for the lack of a single approach. A vital step to achieve this combination is the development of tissue-mimicking phantoms that can satisfy both microwave and ultrasound physical properties at the same time before clinical application. This thesis mainly focused on the breast tissue-mimicking phantom for combined ultrasound and microwave imaging. Four breast tissues (skin, fat, fibroglandular and tumour) were mainly mimicked by canola oil, coconut oil and polyvinylpyrrolidone (PVP) powder with agar and glass beads. First of all, in order to determine the formulae used to mimic different tissues, the interactions between the ratio of ingredients and properties were established by preparing 36 recipes and then measuring both ultrasound and microwave properties. Secondly, after deciding on the formulae to meet the requirements, five box phantoms based on gross breast anatomy with different internal structures had been fabricated and were scanned by a microwave transmission system and an L7-4 ultrasound transducer to obtain microwave and ultrasound images separately. Finally, microwave permittivity maps in horizontal direction were obtained from microwave imaging scan, and 3D images were obtained after segmentation and reconstruction of ultrasound images. The interior design of the five box phantoms could be observed and distinguished by microwave and ultrasound imaging.
Open Access
Deep-learning-based Multi-visit Magnetic Resonance Imaging Reconstruction: Proof of Concept and Robustness Evaluation on a Cohort of Glioblastoma Patients
(2023-01-19) Beauferris, Youssef; Medeiros de Souza, Roberto; Frayne, Richard; Fear, Elise
Magnetic Resonance (MR) imaging is a powerful imaging technique for assessing brain-related diseases. However, MR scans suffer from long acquisition times and as a consequence, patients in Canada must wait extensive periods for access to a scanning session. Compressed Sensing (CS) and Parallel Imaging (PI) are two proven techniques employed to enable accelerated acquisitions. However, they both require complex reconstruction algorithms which disable real-time results. The renewed advent of deep-learning has helped tackle this problem of long reconstruction times. But, currently deep-learning based reconstruction methods do not leverage the wealth of mutual information contained across multiple patient visits to the scanner. This led to the proposal of the Multi-visit Integration Model (MIM) which is a framework for reconstructing a follow-up scan, that has been aggressively undersampled, by leveraging a previous scan. This thesis aims to investigate the performance of the MIM when similarity is not guaranteed between the previous and follow-up scan, such as in the case of glioblastoma patients. The results demonstrated that the MIM leaves localized regions, which have undergone a structural change from one scan to the next, the same. However, this conservative behaviour is not demonstrated during our robustness analysis when synthetic lesions are added to the previous scan to simulate a structural change. The effect of the single-visit reconstruction model on the multi-visit reconstruction performance demonstrated that regardless of the model used, statistically significant improvements to reconstruction quality were observed after multi-visit integration. Multi-visit reconstruction produced using older scans compared to newer scans was found to be of lower quality but still did not introduce biases towards the previous time-point. Finally, the accumulation of system error when using a multi-visit reconstruction as a previous scan in the MIM was minimal. This investigation provided insight into the behaviour of multi-visit integration in the face of structural brain changes and paves the early road towards clinical adoption.
Open Access
Design of a dielectric immersed tapered slotline antenna for second generation tissue sensing adaptive radar system
(2006) Shenouda, Maged Hishmat Hakeem; Fear, Elise
Open Access
Discrete Fourier Transform Techniques to Improve Diagnosis Accuracy in Biomedical Applications
(2018-01-08) Adibpour, Paniz; Smith, Michael; Fear, Elise; Frayne, Richard; Nielsen, John
Transforming acquired data in time or space is necessary for many applications, due to practical constraints on time-domain sampling at high data rates or the requirement for algorithms to process frequency-domain data during the image reconstruction procedure. Therefore, the discrete Fourier transform (DFT) plays an important role in many fields for preprocessing, reconstruction or data analysis stages of algorithms. The hardware or physical constraints also necessitate acquisition of limited length raw data which results in DFT-imposed distortions after data processing for which low pass filters are considered as general solution. Through this thesis, fundamental DFT properties are investigated and an optimization method is introduced to take advantage of these properties. This method is a potential alternative to low pass filters which impose resolution loss to processed data. The formalized method is examined and validated using preliminary observer metrics for two magnetic resonance imaging reconstruction approaches and a microwave imaging technique.
Open Access
Energy Control and Storage to Promote High PV Penetration in Weak Distribution Networks
(2023-07) Rashid, Muhammad Ishaq; Knight, Andrew; Nowicki, Edwin; Galiano, Ignacio; Fear, Elise; Saleh, Saleh
The integration of Photovoltaic Distributed Energy Resources (PVDER) and central energy storage in low voltage distribution networks has been rapidly increasing in popularity within the last decade. Such integration can offer many benefits but may also harm the grid if not managed properly. This research focuses on the technical challenges in weak distribution networks arising from high PVDER penetration. The work concentrates on the issues of feeder voltage regulation, network power factor, generation fair access of participation, system losses and battery energy storage performance in cold weather. The deployment of central battery storage can increase the PV hosting capacity of network. However, battery performance evaluation in Canadian climate is very limited. Practical cycle testing of a vanadium redox flow battery and Li-Ion battery are performed during the winter season in Alberta, Canada. The common methods of Volt/VAr and Volt/Watt used for PVDER local power injection control focus on voltage regulation. However, the side effects impacting the network power factor and generation fair equity are not widely investigated in the literature. This research evaluates the performance of those techniques under real life load demand and PV generation data from a rural distribution network in Alberta, Canada. It becomes evident that under high PV generation scenarios the techniques are inadequate in maintaining an acceptable voltage profiles and severely impact the network power factor and fair generation access. New local control algorithms are developed to mitigate the aforementioned issues. Local PVDER power injection control can be limited in achieving the desired operational outcome. Nonetheless, its use is sometimes the only viable alternative due to the lack or unreliability of communication capability in the distribution network. For when such capability exists, a coordinated algorithm with optimized fair access to generation is developed and evaluated under the same real life demand and PV generation data. The optimized fair access control is able to mitigate the voltage regulation and power factor issues while achieving a high degree of fair generation equity and low system losses.
Open Access
Estimation of Three-Dimensional Breast Features from Standard Two View Mammograms
(2010) Curtis, Charlotte; Fear, Elise; Frayne, Richard
Open Access
Extremely Low Frequency Detection for Biometric Sensing
(2023-09-10) White, Oliver; Yadid-Pecht, Orly; Donev, Jason; Fear, Elise; Curiel, Laura
Human detection is an important aspect of security and surveillance and is a highly researched area, typically using cameras, infrared sensors, or radar. Extremely Low Frequency (ELF) electromagnetic waves, ranging from 0.03 Hz to 300 Hz, offer unique advantages for human detection due to their ability to penetrate various materials, including walls and obstacles. The human body generates a weak magnetic field due to various physiological processes and is measurable within the ELF range. The goal of this thesis is to detect ELF signals emitted by human subjects using an ELF magnetometer and to rapidly distinguish the presence of a human from background noise. A large air-core magnetic induction coil is designed and constructed in conjunction with a low-noise, high-impedance amplification circuit to independently measure the ELF signals from background noise and two human subjects in a semi-remote, non-shielded location. A sampling frequency of 44.1 kHz was used, and the power spectrum density (PSD) was computed for each 3-second data segment. Using a Recurrent Neural Network (RNN) structure, several binary classification models were trained with supervised learning on the PSD sequences between 0.3 and 30 Hz to distinguish background data from human data. The candidate RNN models achieved an average accuracy of above 80% using K-Fold cross-validation on the limited dataset. Overall, the performance of the RNNs was high for one subject, and low for the other subject. It was found that a significant factor in distinguishing a human from background noise was the change in diurnal ELF background power levels between measurements. The results indicate the use of ELF data from humans has the potential for rapid human detection, however, requires a larger dataset for further consideration as a biometric device.