Manual or automatic delineation of the esophageal tumor in CT images is known to be very challenging. This is due to the low contrast between the tumor and adjacent tissues, the anatomical variation of the esophagus, as well as the occasional presence of foreign bodies (e.g. feeding tubes). Physicians therefore usually exploit additional knowledge such as endoscopic findings, clinical history, additional imaging modalities like PET scans. Achieving his additional information is time-consuming, while the results are error-prone and might lead to non-deterministic results. In this paper we aim to investigate if and to what extent a simplified clinical workflow based on CT alone, allows one to automatically segment the esophageal tumor with sufficient quality. For this purpose, we present a fully automatic end-to-end esophageal tumor segmentation method based on convolutional neural networks (CNNs). The proposed network, called Dilated Dense Attention Unet (DDAUnet), leverages spatial and channel attention gates in each dense block to selectively concentrate on determinant feature maps and regions. Dilated convolutional layers are used to manage GPU memory and increase the network receptive field. We collected a dataset of 792 scans from 288 distinct patients including varying anatomies with air pockets, feeding tubes and proximal tumors. Repeatability and reproducibility studies were conducted for three distinct splits of training and validation sets. The proposed network achieved a DSC value of 0.79 ± 0.20, a mean surface distance of 5.4 ± 20.2mm and 95% Hausdorff distance of 14.7 ± 25.0mm for 287 test scans, demonstrating promising results with a simplified clinical workflow based on CT alone. Our code is publicly available via https://github.com/yousefis/DenseUnet_Esophagus_Segmentation .

Although the accuracy of MEMS gyroscopes has been extremely improved, in some aspects, such as stability of bias, they still suffer from some big error sources, like run-to run bias, which determines the sensor price but is not negligible even inexpensive sensors. Due to the fact that run-to-run bias is a kind of stochastic parameter, it has to be measured by utilizing online methods. Utilizing a novel, fast and efficient vision-based rotation estimation algorithm for ground vehicles, we have developed a visual gyroscope that is used in our sensor fusion system, in order to estimate run-to-run bias of the MEMS gyroscope, accurately. Comparing with similar approaches that use GPS, odometer, accelerometer or magnetometer, the most important advantage of the proposed vision-based sensor-fusion framework is its accuracy in accelerated motions. Moreover, it can be used at environments that have a magnetic field, such as urban environments, without depending on external signals. We have evaluated the efficiency of the proposed system using real datasets, recorded from a car moving in urban areas. According to our experimental results, the proposed algorithm is capable of estimating bias of gyroscope after a convergence time of about6 seconds and improving the accuracy of the MEMS gyroscope, which provides the possibility of using cheaper sensors for high accuracy demands.

Basically, charge recovery logic styles have been devised for low-power purposes. However, they have some other characteristics such as inherent pipelining mechanism, low data-dependent power consumption, and low electromagnetic radiations which are usually neglected by researchers. These properties can be useful in other application areas such as side channel attack resistant cryptographic hardware. This paper addresses these properties of charge recovery logics and examines a common one, called 2N-2N2P, as a side channel attack countermeasure by information theoretic evaluation metrics. The observed results show that the usage of this logic style leads to improve DPA-resistance as well as energy saving. Moreover, The behavior of supply currents in simulations reveals that electromagnetic emanation seems to be feeble, compared to other cell level DPA countermeasures.