93 research outputs found
Recommended from our members
Left ventricular geometry during unloading and the end-systolic pressure volume relationship: Measurement with a modified real-time MRI-based method in normal sheep
No full textThe left ventricular (LV) end-systolic (ES) pressure volume relationship (ESPVR) is the cornerstone of systolic LV function analysis. We describe a 2D real-time (RT) MRI-based method (RTPVR) with separate software tools for 1) semi-automatic level set-based shape prior method (LSSPM) of the LV, 2) generation of synchronized pressure area loops and 3) calculation of the ESPVR. We used the RTPVR method to measure ventricular geometry, ES pressure area relationship (ESPAR) and ESPVR during vena cava occlusion (VCO) in normal sheep. 14 adult sheep were anesthetized and underwent measurement of LV systolic function. Ten of the 14 sheep underwent RTMRI and eight of the 14 underwent measurement with conductance catheter; 4 had both RTMRI and conductance measurements. 2D cross sectional RTMRI were performed at apex, mid-ventricle and base levels during separate VCOs. The Dice similarity coefficient was used to compare LSSPM and manual image segmentation and thus determine LSSPM accuracy. LV cross-sectional area, major and minor axis length, axis ratio, major axis orientation angle and ESPAR were measured at each LV level. ESPVR was calculated with a trapezoidal rule. The Dice similarity coefficient between LSSPM and manual segmentation by two readers was 87.31±2.51% and 88.13±3.43%. All cross sections became more elliptical during VCO. The major axis orientation shifted during VCO but remained in the septo-lateral direction. LV chamber obliteration at the apical level occurred during VCO in 7 of 10 sheep that underwent RTMRI. ESPAR was non-linear at all levels. Finally, ESPVR was non-linear because of apical collapse. ESPVR measured by conductance catheter (EES,Index = 2.23±0.66 mmHg/ml/m2) and RT (EES,Index = 2.31±0.31 mmHg/ml/m2) was not significantly different. LSSPM segmentation of 2D RT MRI images is accurate and allows calculation of LV geometry, ESPAR and ESPVR during VCO. In the future, RTPVR will facilitate determination of regional systolic material parameters underlying ESPVR
My Data Body / Your Data Body
No full textSSHRC IDG awarded 2020: Building on research on the use of the medically scanned body to create artistic prints, sculptures and installations, I will work with radiologists and computer scientists to create two art installations, My Data Body, which focuses on representing by own medical and personal data, and Your Data Body, which will focus on representing the data of others. My Data Body will provide viewers with the experience of exploring a full body 3D and 4D magnetic resonance scan dataset in virtual reality (VR).Your Data Body will be made with ‘donated’ datasets of body parts made visible and manipulable in augmented reality (AR), where the virtual overlays on reality. Embedded in the body parts will be a personal story (also 'donated'), that will be audible as long as the scanned body part is being interacted with. Both works will be part of installations that include prints and sculptures generated from the data. My Data Body/Your Data Body will raise many ethical and theoretical questions as well as significant technicaland aesthetic challenges such as; what are the ethics of working artistically my own data and the data of others? What new aesthetic considerations are raised when viewers are able to enter into, pick up and manipulate 3D and 4D scanned bodies? What do these new visions of the digitized body and identity reveal about the human in the digital age? How do they challenge or support existing theoretical frameworks surrounding posthumanism, poststructuralism and technofeminism
Quantum Annealing for Machine Learning: Exploring NISQ Optimization for Image Processing
No full textQuantum computing has a lot of potential for applications of artificial intelligence, even during the current Noisy Intermediate-Scale Quantum (NISQ) era. One important application in image processing is image denoising. Existing methods use either a simple quantum method on binary images or various forms of hybrid quantumclassical methods. In this work, we extend a fully quantum denoising method based on random Markov fields for non-binary images and demonstrate its effectiveness. Our results show that solving such problems with full-depth images using NISQ anneals remains computationally challenging, with limited accuracy due to hardware constraints and noise.
Quantum anneals are known for their ability to efficiently approximate solutions, leveraging unique quantum phenomena such as quantum tunneling to navigate complex optimization landscapes. While these devices have been employed in various NP-hard problems that are central to artificial intelligence, the extent to which quantum tunneling contributes to performance enhancements is still unclear. To address this, we perform an experimental analysis examining the relationship between the complexity of the optimization energy landscape and the performance of quantum annealing. Our findings provide insights into the capabilities and limitations of current quantum anneals for solving complex optimization problems
Optimized U-Net for Left Ventricle Segmentation
No full textThe left ventricle segmentation is an important medical imaging task necessary to measure a patient's heart pumping efficiency. Recently, convolutional neural networks (CNN) have shown great potential in achieving state-of-the-art segmentation for such applications. However, most of the research is focusing on building complicated variations of these networks with modest changes to its performance. There is little to no insights on how these CNNs work and most of them are unfortunately treated the neural network as a black box. In this thesis, the famous U-Net architecture is used to segment the left ventricle from cardiac magnetic resonance (MR) images because of its simplicity and ability to analyze images at multiple scales. Posterior analysis of the network functionality demonstrates that by replacing the first set of layers of the U-Net with fixed filters, there is little change in performance compared to its fully connected version. This optimization was achieved by performing a Fourier analysis and visualization of the convolution layers after the completion of the network training phase. This analysis allows us to discover that some early layers approximate uniform filters which can then be replaced by fixed uniform kernel weights. Furthermore, in a separate experiment by removing the middle layers of the U-Net one can reduce the number of U-Net parameters from 31 million to 0.5 million to achieve faster prediction time without compromising the performance. Experimental results and analysis are presented
Parallel implementation of a nonrigid image registration algorithm for lung tumor boundary tracking in quasi real-time MRI
No full text
- …
