Search for: All records

This paper describes a novel framework for executing a network of trained deep neural network (DNN) models on commercial-off-the-shelf devices that are deployed in an IoT environment. The scenario consists of two devices connected by a wireless network: a user-end device (U), which is a low-end, energy and performance-limited processor, and a cloudlet (C), which is a substantially higher performance and energy-unconstrained processor. The goal is to distribute the computation of the DNN models between U and C to minimize the energy consumption of U while taking into account the variability in the wireless channel delay and the performance overhead of executing models in parallel. The proposed framework was implemented using an NVIDIA Jetson Nano for U and a Dell workstation with Titan Xp GPU as C. Experiments demonstrate significant improvements both in terms of energy consumption of U and processing delay. 

PurposeThe goal of this study was to determine the accuracy of displacement‐encoding with stimulated echoes (DENSE) MRI in a tissue motion phantom with displacements representative of those observed in human brain tissue. MethodsThe phantom was comprised of a plastic shaft rotated at a constant speed. The rotational motion was converted to a vertical displacement through a camshaft. The phantom generated repeatable cyclical displacement waveforms with a peak displacement ranging from 92 µm to 1.04 mm at 1‐Hz frequency. The surface displacement of the tissue was obtained using a laser Doppler vibrometer (LDV) before and after the DENSE MRI scans to check for repeatability. The accuracy of DENSE MRI displacement was assessed by comparing the laser Doppler vibrometer and DENSE MRI waveforms. ResultsLaser Doppler vibrometer measurements of the tissue motion demonstrated excellent cycle‐to‐cycle repeatability with a maximum root mean square error of 9 µm between the ensemble‐averaged displacement waveform and the individual waveforms over 180 cycles. The maximum difference between DENSE MRI and the laser Doppler vibrometer waveforms ranged from 15 to 50 µm. Additionally, the peak‐to‐peak difference between the 2 waveforms ranged from 1 to 18 µm. ConclusionUsing a tissue phantom undergoing cyclical motion, we demonstrated the percent accuracy of DENSE MRI to measure displacement similar to that observed for in vivo cardiac‐induced brain tissue.