No abstract

PurposeThe aim of this study was to compare the effective dose for imaging of mandible between multi-detector computed tomography (MDCT) and cone-beam computed tomography (CBCT). An MDCT with low dose technique was also compared with them.Materials and MethodsThermoluminescent dosimeter (TLD) chips were placed at 25 organ sites of an anthropomorphic phantom. The mandible of the phantom was exposed using 2 different types of MDCT units (Somatom Sensation 10 for standard-dose MDCT, Somatom Emotion 6 for low-dose MDCT) and 3 different CBCT units (AZ3000CT, Implagraphy, and Kavo 3D eXaM). The radiation absorbed dose was measured and the effective dose was calculated according to the ICRP 2007 report.ResultsThe effective dose was the highest for Somatom Sensation 10 (425.84 µSv), followed by AZ3000CT (332.4 µSv), Somatom Emotion 6 (199.38 µSv), and 3D eXaM (111.6 µSv); it was the lowest for Implagraphy (83.09 µSv). The CBCT showed significant variation in dose level with different device.ConclusionThe effective doses of MDCTs were not significantly different from those of CBCTs for imaging of mandible. The effective dose of MDCT could be markedly decreased by using the low-dose technique.

Purpose: This study was performed to investigate the effects of energy level, reconstruction kernel, and tube rotation time on Hounsfield unit (HU) values of hydroxyapatite (HA) in virtual monochromatic images (VMIs) obtained with dual-energy computed tomography (DECT) (Siemens Healthineers, Erlangen, Germany). Materials and Methods: A bone density calibration phantom with 3 HA inserts of different densities (CTWATER ® ; 0, 100, and 200 mg of HA/cm 3) was scanned using a twin-beam DECT scanner at 120 kVp with tube rotation times of 0.5 and 1.0 seconds. The VMIs were reconstructed by changing the energy level (with options of 40 keV, 70 keV, and 140 keV). In order to investigate the impact of the reconstruction kernel, virtual monochromatic images were reconstructed after changing the kernel from body regular 40 (Br40) to head regular 40 (Hr40) in the reconstruction phase. The mean HU value was measured by placing a circular region of interests (ROIs) in the middle of each insert obtained from the VMIs. The HU values were compared with regard to energy level, reconstruction kernel, and tube rotation time. results: Hydroxyapatite density was strongly correlated with HU values (correlation coefficient = 0.678, P<0.05). For the HA 100 and 200 inserts, HU decreased significantly at increased energy levels (correlation coefficient =-0.538, P<0.05) but increased by 70 HU when using Hr40 rather than Br40 (correlation coefficient = 0.158, P<0.05). The tube rotation time did not significantly affect the HU (P>0.05). Conclusion: The HU values of hydroxyapatite were strongly correlated with hydroxyapatite density and energy level in VMIs obtained with DECT.

Cybersickness is a symptom of dizziness that occurs while experiencing Virtual Reality (VR) technology and it is presumed to occur mainly by crosstalk between the sensory and cognitive systems. However, since the sensory and cognitive systems cannot be measured objectively, it is difficult to measure cybersickness. Therefore, methodologies for measuring cybersickness have been studied in various ways. Traditional studies have collected answers to questionnaires or analyzed EEG data using machine learning algorithms. However, the system relying on the questionnaires lacks objectivity, and it is difficult to obtain highly accurate measurements with the machine learning algorithms. In this work, we apply Deep Neural Network (DNN) deep learning algorithm for objective cybersickness measurement from EEG data. We also propose a data preprocessing for learning and network structures allowing us to achieve high performance when learning EEG data with the deep learning algorithms. Our approach provides cybersickness measurement with an accuracy up to 98.88%. Besides, we analyze video characteristics where cybersickness occurs by examining the video segments causing cybersickness in the experiments. We discover that cybersickness happens even in unusually persistent changes in the darkness such as the light in a room keeps switching on and off.

For various target tracking applications, it is well known that the Kalman filter is the optimal estimator(in the minimum mean-square sense) to predict and estimate the state(position and/or velocity) of linear dynamical systems driven by Gaussian stochastic noise. In the case of nonlinear systems, Extended Kalman filter(EKF) and/or Unscented Kalman filter(UKF) are widely used, which can be viewed as approximations of the(linear) Kalman filter in the sense of the conditional expectation. However, to implement EKF and UKF, the exact dynamical model information and the statistical information of noise are still required. In this paper, we propose the recurrent neural-network based Kalman filter, where its Kalman gain is obtained via the proposed GRU-LSTM based neural-network framework that does not need the precise model information as well as the noise covariance information. By the proposed neural-network based Kalman filter, the state estimation performance is enhanced in terms of the tracking error, which is verified through various linear and nonlinear tracking problems with incomplete model and statistical covariance information.