998 research outputs found
Fractionated muscle activity in gait initiation of Parkinson’s participants may be exacerbated by REM sleep without atonia
No full textThis research was supported by the Undergraduate Research Opportunities Program (UROP).Kim, Minwoo. (2020). Fractionated muscle activity in gait initiation of Parkinson’s participants may be exacerbated by REM sleep without atonia. Retrieved from the University Digital Conservancy, https://hdl.handle.net/11299/216185
Ultrasound tissue perfusion imaging
Get PDFEnhanced blood perfusion in a tissue mass is an indication of neo-vascularity and potential malignancy. Ultrasonic pulsed Doppler imaging is a safe and economical modality for noninvasive monitoring of blood flow. However, weak blood echoes make it difficult to detect perfusion using standard methods without the expense of contrast enhancement. Additionally, imaging requires high sensitivity to slow, disorganized blood-flow patterns while simultaneously rejecting clutter and noise. An approach to address these challenges involves arranging acquisition data in a multi-dimensional structure to facilitate the characterization and separation of independent scattering sources. The resulting data array involves a linear combination of spatial, slow-time (kHz-order sampling), and frame-time (Hz-order sampling) coordinates. Applying an eigenfilter that exploits higher-order singular value decomposition (HOSVD) can technically transform the array and reduce the dimensions to yield power estimates for blood flow and perfusion that are well isolated from tissue clutter. Studies using microcirculation-mimicking simulations and phantoms enable the optimization of the filtering algorithm to maximize estimation efficiency. These techniques are applied to murine models of ischemia and melanoma at 24 MHz to form perfusion images. The results show enhancements of tissue perfusion maps, which help researchers access lesions without contrast enhancement. In a study aimed at peripheral artery disease (PAD), the enhanced sensitivity and specificity of ultrasonic-pulsed-Doppler imaging enable differentiation of perfusion between healthy and ischemic states. In addition, the use of the new ultrasound imaging coupled with other imaging modalities helps to illuminate the complex mechanism that mediates neovascularization in response to vascular occlusion. Consequently, these techniques have the potential to increase the effectiveness of existing medical imaging technologies in safe, cost-effective ways that promote sustainable medicine.Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2018-09-27 without embargo termsThe student, Minwoo Kim, accepted the attached license on 2018-04-25 at 15:51.The student, Minwoo Kim, submitted this Dissertation for approval on 2018-04-25 at 16:11.This Dissertation was approved for publication on 2018-04-26 at 14:28.DSpace SAF Submission Ingestion Package generated from Vireo submission #12493 on 2018-09-27 at 10:43:32Made available in DSpace on 2018-09-27T16:11:07Z (GMT). No. of bitstreams: 3
KIM-DISSERTATION-2018.pdf: 9443822 bytes, checksum: 7be1a406d0a6bcdfb0ccae969d4e1eff (MD5)
LICENSE.txt: 4207 bytes, checksum: e6123bab2dd4ad2eb100e0120b40eac5 (MD5)
PROQUEST_LICENSE.txt: 4553 bytes, checksum: f70643ca8f8e0dd594708e6ae28a2857 (MD5)
Previous issue date: 2018-04-2
Measurements of self-excited instabilities and nitrogen oxides emissions in a multi-element lean-premixed hydrogen/methane/air flame ensemble
No full textMeasurements of self-excited instabilities and nitrogen oxides emissions in a multi-element lean-premixed hydrogen/methane/air flame ensemble
No full textUnderstanding the distinguishing physical properties of multi-element lean-premixed high hydrogen content flames is expected to be integral to the development of carbon-neutral, and ultimately carbon-free, gas turbine combustion systems. Despite their fundamental importance, the thermoacoustic and emission-related characteristics of such small-scale flame ensembles are not thoroughly understood, particularly for the full range of 0 to 100% hydrogen content blended with methane fuel. Here we investigate the structure and collective behavior of a multi-element lean-premixed hydrogen/methane/air flame ensemble using measurements of nitrogen oxides emissions and self-excited instability, combined with OH* and OH PLIF flame visualizations. Our results indicate that the system's responses can be classified into several distinctive stages according to their static and dynamic stability, including flame blowoff and thermoacoustically stable regions under relatively low hydrogen concentration conditions, low-frequency self-excited instabilities in intermediate hydrogen concentration, and triggering of intense pressure perturbations at about 1.7 kHz under high- or pure hydrogen combustion conditions. While the low-frequency combustion dynamics are dominated by axisymmetric translational movements of parallel flame fronts, the higher frequency response originates from significant lateral modulations accompanied by small-scale vortical rollup and flame surface annihilation due to front merging and pinch-off. Longitudinal-to-transverse dynamic transition is observed to play a mechanistic role in kinematically accommodating higher-frequency heat release rate fluctuations, and this newly identified mechanism suggests the possibility of high-frequency transverse modes, if such lateral motions are strong enough to induce inter-element flame interactions. In contrast to the substantial differences in thermoacoustic properties for different fuel compositions, the total nitrogen oxides emissions are found to depend primarily on adiabatic flame temperature; the influence of fuel composition is limited to approximately 20% under the inlet conditions considered.
sj-docx-1-eso-10.1177_23969873221144814 – Supplemental material for Effects of prior antiplatelet use on futile reperfusion in patients with acute ischemic stroke receiving endovascular treatment
No full textSupplemental material, sj-docx-1-eso-10.1177_23969873221144814 for Effects of prior antiplatelet use on futile reperfusion in patients with acute ischemic stroke receiving endovascular treatment by Jong-Hee Sohn, Chulho Kim, Minwoo Lee, Yerim Kim, Hee Jung Mo, Kyung-Ho Yu and Sang-Hwa Lee in European Stroke Journal</p
Target Classification and Prediction of Unguided Rocket Trajectories Using Deep Neural Networks
No full textTowards monocular vision-based autonomous flight through deep reinforcement learning
Get PDFThis paper proposes an obstacle avoidance strategy for small multi-rotor drones with a monocular camera using deep reinforcement learning. The proposed method is composed of two steps: depth estimation and navigation decision making. For the depth estimation step, a pre-trained depth estimation algorithm based on the convolutional neural network is used. On the navigation decision making step, a dueling double deep Q-network is employed with a well-designed reward function. The network is trained using the robot operating system and Gazebo simulation environment. To validate the performance and robustness of the proposed approach, simulations and real experiments have been carried out using a Parrot Bebop2 drone in various complex indoor environments. We demonstrate that the proposed algorithm successfully travels along the narrow corridors with the texture free walls, people, and boxes
Parametric study on the flight envelope of a radio-frequency ion thruster based atmosphere-breathing electric propulsion system
No full textThe atmosphere-breathing electric propulsion (ABEP) system utilizes atmospheric species as propellants to generate thrust for drag compensation of a satellite in very-low-Earth-orbit (VLEO). A parametric approach is used to assess the impact of different parameters on the flight envelope of a sample VLEO satellite with an ABEP system based on a radio-frequency ion thruster (RIT). The considered parameters include capture efficiency, maximum input power, solar activity, and atomic oxygen recombination factor. The NRLMSIS 2.0 atmosphere model is used to determine the flow conditions at VLEO, at a target altitude of 150–300 km with high, moderate, and low solar activity levels. DSMC method is employed to calculate the drag of the sample satellite with the ABEP system. The 0-D model of the RIT discharge chamber is used to predict the thrust of the RIT-based ABEP system. The flight envelope of a sample RIT-based ABEP system shows that drag can be compensated at altitudes between 196 km and 248 km. Increasing capture efficiency and maximum input power expands the feasible range for drag compensation to higher and lower altitudes, respectively. Also, the flight envelope shifts to higher altitudes with increasing solar activity levels. However, the atomic oxygen recombination factor of the intake device has minimal effects on the flight envelope.
Monocular vision-based time-to-collision estimation for small drones by domain adaptation of simulated images
No full textRecently, there is an increasing demand for small drones owing to their small size and agility in complex indoor environments. Accordingly, safety issues for navigating small drones become of significant importance. For drones to be able to navigate safely through complex environments, it would be useful to estimate accurate time-to-collision (TTC) to obstacles. To this end, in this paper, we propose a deep learning-based TTC estimation algorithm. To train generalizable neural networks for TTC estimation, large datasets including collision cases are needed. However, in real-world environments, it is impractical and infeasible to collide drones with obstacles to collect a significant amount of data. Simulation environments could facilitate the data acquisition procedure, but the data from simulated environments could be quite different from those of real environments, which is commonly termed as reality-gap. In this study, to reduce this reality-gap, sim-to-real methods based on a variant of the generative adversarial network are used to convert simulated images into real world-like synthetic images. Besides, to consider the uncertainties that come from using the synthetic dataset, the aleatoric loss function and Monte Carlo dropout method are employed. Furthermore, we improve the performance of the deep learning-based TTC estimation algorithm by replacing conventional convolutional neural networks (CNNs) with convolutional long short-term memory (ConvLSTM) layers which are known to be better at handling time-series data than CNNs. To validate the performance of the proposed framework, real flight experiments have been carried out in various indoor environments. Our proposed framework decreases the average TTC estimation error by 0.21 s compared with the baseline approach with CNNs
A new controller design method for an electric power steering system based on a target steering torque feedback controller
No full textThis paper describes a novel electric power steering (EPS) system controller that distinguishes the steering feel design problem from the system stability and control performance problem and provides a design methodology for each problem. The suggested controller, called the target steering torque feedback controller, consists of two modules: the steering feel generation logic module defines the target steering torque that the driver should feel under a given driving condition, and the steering torque feedback controller module controls the assist motor so that the driver actually feels the defined target torque. By disassociating all nonlinear elements related to steering feel tuning from the feedback controller, the inherent difficulties of the conventional EPS controller in terms of the coupling of system stability analysis and steering feel design are overcome. Rack-force-based steering feel generation logic that creates a familiar steering feel for drivers while transmitting the road condition information is suggested and design criteria of the steering torque feedback controller in the frequency domain are proposed to ensure robust stability, tracking performance and noise attenuation. In addition, a disturbance observer is applied to the feedback controller to compensate for the effect of disturbances and to improve tracking performance. The proposed controller design method is verified by conducting computer simulations and a vehicle test.
- …
