Sharif Digital Repository

In recent years, deep learning-based networks have achieved state-of-the-art performance in medical image segmentation. Among the existing networks, U-Net has been successfully applied on medical image segmentation. In this paper, we propose an extension of U-Net, Bi-directional ConvLSTM U-Net with Densely connected convolutions (BCDU-Net), for medical image segmentation, in which we take full advantages of U-Net, bi-directional ConvLSTM (BConvLSTM) and the mechanism of dense convolutions. Instead of a simple concatenation in the skip connection of U-Net, we employ BConvLSTM to combine the feature maps extracted from the corresponding encoding path and the previous decoding up-convolutional... 

In the open-pit mines and civil projects, drilling and blasting is the most common method for rock fragmentation aims. This article proposes a new hybrid forecasting model based on firefly algorithm, as an algorithm optimizer, combined with the adaptive neuro-fuzzy inference system for estimating the fragmentation. In this regard, 72 datasets were collected from Shur river dam region, and the required parameters were measured. Using the different input parameters, six hybrid models were constructed. In these models, 58 and 14 data were used for training and testing, respectively. The proposed hybrid models were then evaluated in accordance with statistical criteria such as coefficient of... 

Reliability evaluations play a significant role in engineering applications to ensure the serviceability and safety of advanced structures such as those made of composites. Here, a dynamic reliability evaluation analysis based on the probability density evolution Method (PDEM) has been adapted to assess the reliability of composite structures under uncertainties within the material properties and the external loadings. A Back-Propagation Neural Network approach is employed to identify the system's nonlinear structural response, which is often the case under large deformations. To exemplify, a split Hopkinson pressure bar system was employed to mimic the mechanical behavior of a... 

The main focus of this paper is to present thermodynamic and economic analyses and multi-objective optimization of a novel geothermal-solar multigeneration system. The system aims to produce hydrogen, freshwater, electricity, cooling load, and hot water and designed based on geothermal and solar energy. After modeling and thermodynamic and economic analysis, exergy destruction rate, exergy efficiency and, cost rate were calculated for each component of the system. The results showed that the highest amount of exergy destruction was related to parabolic trough collectors (PTCs) and absorption chillers. To select the geothermal fluid of the organic Rankine cycle (ORC), several different fluids... 

Anatomical image segmentation is one of the foundations for medical planning. Recently, convolutional neural networks (CNN) have achieved much success in segmenting volumetric (3D) images when a large number of fully annotated 3D samples are available. However, rarely a volumetric medical image dataset containing a sufficient number of segmented 3D images is accessible since providing manual segmentation masks is monotonous and time-consuming. Thus, to alleviate the burden of manual annotation, we attempt to effectively train a 3D CNN using a sparse annotation where ground truth on just one 2D slice of the axial axis of each training 3D image is available. To tackle this problem, we propose... 

Predictive business process monitoring is concerned with predicting the process-related Key Performance Indicators (KPIs) and forecasting the future behavior of the process in realtime. Despite the amount of work contributed by researches to this field of research, the performance of existing solutions is not desirable for practical settings. Indeed, these approaches are typically context-unaware and lack generality. However, in real-life use cases, business processes are not isolated from the surrounding working environment, and thus they are influenced by many contextual events, such as events generated by IoT devices. To the best of our knowledge, there is no comprehensive study... 

The artificial neural network (ANN) is a sort of machine learning method which has been used in determination of risk of human disorders. In the current investigation, we have created an ANN and trained it based on the genetic data of 401 multiple sclerosis (MS) patients and 390 healthy subjects. Single nucleotide polymorphisms (SNPs) within ANRIL (rs1333045, rs1333048, rs4977574 and rs10757278), EVI5 (rs6680578, rs10735781 and rs11810217), ACE (rs4359 and rs1799752), MALAT1 (rs619586 and rs3200401), GAS5 (rs2067079 and rs6790), H19 (rs2839698 and rs217727), NINJ2 (rs11833579 and rs3809263), GRM7 (rs6782011 and rs779867), VLA4 (rs1143676), CBLB (rs12487066) and VEGFA (rs3025039 and... 

Spinal posture including thorax/pelvis orientations as well as loads on the intervertebral discs are crucial parameters in biomechanical models and ergonomics to evaluate the risk of low back injury. In vivo measurement of spinal posture toward estimation of spine loads requires the common motion capture techniques and laboratory instruments that are costly and time-consuming. Hence, a closed loop algorithm including an artificial neural network (ANN) and fuzzy logic is proposed here to predict the L5–S1 segment loads and thorax/pelvis orientations in various 3D reaching activities. Two parts namely a fuzzy logic strategy and an ANN from this algorithm; the former, developed based on the... 

We propose a lightweight scheme where the formation of a data block is changed in such a way that it can tolerate soft errors significantly better than the baseline. The key insight behind our work is that CNN weights are normalized between -1 and 1 after each convolutional layer, and this leaves one bit unused in half-precision floating-point representation. By taking advantage of the unused bit, we create a backup for the most significant bit to protect it against the soft errors. Also, considering the fact that in MLC STT-RAMs the cost of memory operations (read and write), and reliability of a cell are content-dependent (some patterns take larger current and longer time, while they are... 

In the current study, CuO nanoparticles were dispersed in a mixture of Ethylene Glycol-Water (60/40 wt. %) to prepare stable nanofluid in different concentrations (0.05 − 0.8 vol. %). The samples were used as the coolant fluid in a specific car radiator to evaluate the thermal performance of nanofluid and base fluid in the system. Five different and novel Machine-learning methods were applied over experimental data to predict the Nusselt number and output temperature of the coolant in the system. These methods are M5 tree regression, Linear and Cubic Multi-Variate Adaptive Regression Splines (MARS), Radial Basis Function (RBF), and Artificial Neural Network-Levenberg Marquardt Algorithm... 

Computed tomography (CT) simulators are versatile tools for scanning protocol evaluation, optimization of geometrical design parameters, assessment of image reconstruction algorithms, and evaluation of the impact of future innovations attempting to improve the performance of CT scanners. Computational human phantoms (CHPs) play a key role in simulators for the radiation dosimetry and assessment of image quality tasks in the medical x-ray systems. Since the construction of patient-specific CHPs can be both difficult and time-consuming, nominal standard/reference CHPs have been established, yielding significant discrepancies in the special design and optimization demands of patient dose and... 

This paper presents a data-driven approach that utilizes the gathered experimental data to model and control a test rig constructed for the high-powered gearboxes. For simulating a wide variety of operational conditions, the test rig should be capable of providing different speeds and torques; this is possible using a torque-applying system. For this purpose, Electro-Hydraulic Actuators (EHAs) are used. Since applying accurate torque is a crucial demand as it affects the performance evaluation of the gearboxes, precise modelling of the actuation system along with a high-performance controller are required. In order to eliminate the need for to solve complex nonlinear equations of EHA that... 

Elemental analysis can be performed using obtained gamma-ray spectrum of the sample under study. In this work, simple Multi-Layer Perceptron (MLP) neural network models are proposed for analyzing a gamma-ray emitting sample using whole information of its obtained gamma-ray spectrum. Elemental analysis is performed in two fields of study using 3 × 3 inch NaI(Tl) detectors: Radio-Isotope Identification (RIID) and Prompt Gamma Neutron Activation Analysis (PGNAA). The gamma-ray point sources are used for an empirical study in RIID field, while a Monte Carlo simulation study is considered for determining chlorine and water content of crude oil using combination of PGNAA technique and a MLP model.... 

Low back pain (LBP), the leading cause of disability worldwide, remains one of the most common and challenging problems in occupational musculoskeletal disorders. The effective assessment of LBP injury risk, and the design of appropriate treatment modalities and rehabilitation protocols, require accurate estimation of the mechanical spinal loads during different activities. This study aimed to: (1) develop a novel 2D beam-column finite element control-based model of the lumbar spine and compare its predictions for muscle forces and spinal loads to those resulting from a geometrically matched equilibrium-based model; (2) test, using the foregoing control-based finite element model, the... 

Sophisticated surgeons are widely indicating the use of surgical robots in order to reject human error, increase precision, and speed. Among well-known robotic mechanisms, parallel robots are broadly more investigated regarding their special characters as higher acceleration, speed, and accuracy, and less weight. Specific surgical procedures confine, and restrict their workspace, while controlling and validating the robots are complicated regarding to their complex dynamic. To this end, in this paper, a 6-DOF robot, with rotary manipulators, is designed and controlled. Addressing nonlinearity of parallel robots, an innovative methodology is formulated to robustly penalize the error of... 

Tight unmanned aerial vehicle (UAV) autonomous missions such as formation flight and aerial refueling (AR) requires an active controller that works in conjunction with a precise vision-based sensor that is able to extract In-front aircraft relative position and orientation from captured images. A key point in implementing such a sensor is its robustness in the presence of noises and other uncertainties. In this paper, a new vision-based algorithm that uses neural networks to estimate the In-front aircraft relative orientation and position is developed. The accuracy and robustness of the proposed algorithm has been validated via a detailed modeling and a complete virtual environment based on... 

Fractional order PID controllers are suitable for almost all types of dynamic models. In this paper, a new adaptive fractional order PID controller using neural networks is introduced.The overall performance using the proposed adaptive fractional order PID controller is demonstrated through some examples. It is shown that the new controller scheme can give excellent performance and more robustness in comparison with the conventional controllers like GPC  

Abstract- Benefits of PSS application for electromechanical oscillations damping is clarified. A major difficulty of PSS application is its tuning such that its damping characteristics do not fall into overcompensation facing continuously changing power systems. This paper presents an approach for wide area monitoring and control based PSS tuning. The presented approach originated from adaptive tuning based on real time power system state estimation. © 2008 IEEE  

Many complex networks such as social networks and computer show modular structures, where edges between nodes are much denser within modules than between modules. It is strongly believed that cellular networks are also modular, reflecting the relative independence and coherence of different functional units in a cell. In this paper we used a human curated dataset. In this paper we consider each module in the PPI network as ontology. Using techniques in ontology alignment, we compare each pair of modules in the network. We want to see that is there a correlation between the structure of each module or they have totally different structures. Our results show that there is no correlation... 

The problem of time optimal magnetic attitude control is treated and an open loop solution is first obtained using a variational approach. In order to close the control loop, a neural network with time varying weights is proposed as a feedback optimal controller applicable to the time varying nonlinear system. The good robustness and low real-time computational burden of the proposed neuro-controller makes the controller more useful compared to the other control methods. © 2008 IEEE