Artificial neural network approach for prediction of stress–strain curve of near b titanium alloy

In the present study, artificial neural network(ANN) approach was used to predict the stress–strain curve of near beta titanium alloy as a function of volume fractions of a and b. This approach is to develop the best possible combination or neural network(NN) to predict the stress–strain curve. In order to achieve this, three different NN architectures(feed-forward back-propagation network,cascade-forward back-propagation network, and layer recurrent network), three different transfer functions(purelin, Log-Sigmoid, and Tan-Sigmoid), number of hidden layers(1 and 2), number of neurons in the hidden layer(s),and different training algorithms were employed. ANN training modules, the load in terms of strain, and volume fraction of a are the inputs and the stress as an output.ANN system was trained using the prepared training set(a,16 % a, 40 % a, and b stress–strain curves). After training process, test data were used to check system accuracy. It is observed that feed-forward back-propagation network is the fastest, and Log-Sigmoid transfer function is giving the best results. Finally, layer recurrent NN with a single hidden layer consists of 11 neurons, and Log-Sigmoid transfer function using trainlm as training algorithm is giving good result, and average relative error is1.27 ± 1.45 %. In two hidden layers, layer recurrent NN consists of 7 neurons in each hidden layer with trainrp as the training algorithm having the transfer function of LogSigmoid which gives better results. As a result, the NN is founded successful for the prediction of stress–strain curve of near b titanium alloy.     

A Constraint-Elimination Technique for Linearly Constrained Array Processing

A new technique is presented for the problem of linearly constrained multi-channel-array processing. The purpose of the array processing is to reject undesired noise, in the minimum-mean-squared-error sense, while responding to any signal coming from a particular direction with a preset, constrained frequency response. In the constraint-elimination technique proposed in this paper, the original constrained-array-processing problem is transformed to the simple unconstrained-Wiener-filtering problem. Transformation of the problem is achieved by introducing a ?compensating? channel to the multichannel processor. Input to the compensating channel is the channel-averaged input. The constraint is then eliminated by expressing the filter weights of the compensating channel in terms of those of the original processor. An adaptive algorithm is derived by applying the stochastic approximation to the simplified Wiener-filtering problem. Simulation experiments verify that the constrained-array-processing problem is properly transformed to the unconstrained one by the proposed technique. Additional experiments show that the adaptive algorithm generally reduces output power while maintaining the constrained frequency response to the desired signal.     

Interventional radiology with laser in bone and joint

Laser energy is able to ablate, coagulate, and vaporize tissues. Its transmissibility in thin optical fibers makes it an ideal tool for use in percutaneous procedures. This article describes two applications in interventional musculoskeletal radiology. In percutaneous laser disc decompression the laser source is used to vaporize a small portion of the nucleus pulposus. In interstitial laser photocoagulation of osteoid osteoma the laser energy is used to coagulate and destroy the tumor by direct heating.     

The characteristics of electrically short, umbrella top-loaded antennas

Experimental measurements of the electrical characteristics of umbrella top-loaded, electrically short antennas have been performed. Electrically short antennas can be represented by a series RLC circuit with good accuracy up to and including their first series resonant frequency. Electrically short antennas generally have very high Q's and, consequently, narrow bandwidths. Top-loading, such as umbrella top-loading of electrically short monopole antennas, leads to antennas which have lower Q's and larger bandwidths. The experimental measurements were made using a scale model facility. The antennas measured on the scale model facility were one-hundredth scale models of the anticipated full scale antennas. The equivalent inductance, capacitance, resistance, and effective height of these scale model antennas were determined and compared with values of static capacitance, static effective height, and static resistance obtained from a computer calculation [1], [2] and good correlation was obtained. Additionally, the dynamic characteristics of the antennas were determined from the model studies. The dynamic parameters determined were the antenna inductance, the dynamic radiation resistance, and the dynamic effective height. The measured data obtained from the model study are summarized in graphs and tables and compared with other reported experimental results as well as with computer calculations. Nomograms summarizing the computed data have been constructed and are presented.     

A patient-mounted robotic platform for CT-scan guided procedures

In this paper, we present a novel robotic assistant dedicated to medical interventions under computed tomography scan guidance. This compact and lightweight patient-mounted robot is designed so as to fulfill the requirements of most interventional radiology procedures. It is built from an original 5 DOF parallel structure with a semispherical workspace, particularly well suited to CT-scan interventional procedures. The specifications, the design, and the choice of compatible technological solutions are detailed. A preclinical evaluation is presented, with the registration of the robot in the CT-scan.     

Technic of overlapping sphincter anal repair in the treatment of traumatic anal incontinence

Faecal incontinence is an important disabling symptom in the affected patients. Classically, we divide faecal incontinence in two main types: neurogenic faecal incontinence and traumatic anal incontinence. Traumatic anal incontinence is due to causes damaging sphincteric mechanism directly. The aim of the present study was to evaluate the outcome of overlapping sphincter anal repair procedure in the management of traumatic anal incontinence. To this end we studied 27 patients with traumatic anal incontinence who underwent an overlapping sphincter anal repair procedure according to the method described by Parks and McPartlin in 1971. Mean follow up was up three years and was based mainly both on clinical evaluation with anorectal exploration and manometric values carried out on a 6 monthly basis. When the outcome was evaluated in terms of faecal continence our date were similar to those reported by Parks and Fang. In the subjects studied we haven't reported any major complications apart from one case of abscess, one case of wound's infection and one case of stenosis which were efficaciously treated. Our findings supported the view that overlapping sphincter anal repair procedure is the surgical approach of choice in the patients with traumatic anal incontinence.     

Perceived demands and musculoskeletal symptoms among employees of an Iranian petrochemical industry

As a part of a comprehensive ergonomics program, this study was conducted among employees of an Iranian petrochemical industry to determine the prevalence of musculoskeletal symptoms and to examine the relationship between perceived demands and reported symptoms. In this cross-sectional study, 928 randomly selected employees, corresponding to nearly 40% of all employees participated. Nordic Musculoskeletal Disorder Questionnaire and Job Content Questionnaire were used as collecting data tools. The results showed that 73% of the study population had experienced some form of symptoms from the musculoskeletal system during the last 12 months. Knees and lower back symptoms were the most prevalent problem among the employees studied. The results revealed that perceived physical demands were significantly associated with musculoskeletal symptoms (OR ranged from 1.45 to 2.33). Among the perceived physical demands, awkward working postures were most frequently associated with reported musculoskeletal symptoms. Association was also found between perceived psychological demands and reported symptoms. Conflicting demands, waiting on work from other people or departments, interruption that other make, working very fast and time pressure were psychological factors retained in the regression models with OR ≥ 1.49. Based on the findings, it could be concluded that any interventional program for preventing or reducing musculoskeletal symptoms among the petrochemical employees studied had to focus on reducing physical demands, particularly awkward working postures as well as psychological aspect of working environment.

Relevance to industry

In petrochemical industry where employees are involved in both static and dynamic activities, determination of musculoskeletal symptoms contributing factors can be considered as a basis for planning and implementing interventional ergonomics program for preventing musculoskeletal symptoms and improving working conditions.     

Artificial neural network approach for prediction of stress–strain curve of near β titanium alloy

In the present study, artificial neural network（ANN） approach was used to predict the stress-strain curve of near beta titanium alloy as a function of volume fractions of a and b. This approach is to develop the best possible combination or neural network（NN） to predict the stress-strain curve. In order to achieve this, three different NN architectures（feed-forward back-propagation network,cascade-forward back-propagation network, and layer recurrent network）, three different transfer functions（purelin, Log-Sigmoid, and Tan-Sigmoid）, number of hidden layers（1 and 2）, number of neurons in the hidden layer（s）,and different training algorithms were employed. ANN training modules, the load in terms of strain, and volume fraction of a are the inputs and the stress as an output.ANN system was trained using the prepared training set（a,16 % a, 40 % a, and b stress-strain curves）. After training process, test data were used to check system accuracy. It is observed that feed-forward back-propagation network is the fastest, and Log-Sigmoid transfer function is giving the best results. Finally, layer recurrent NN with a single hidden layer consists of 11 neurons, and Log-Sigmoid transfer function using trainlm as training algorithm is giving good result, and average relative error is1.27 ± 1.45 %. In two hidden layers, layer recurrent NN consists of 7 neurons in each hidden layer with trainrp as the training algorithm having the transfer function of LogSigmoid which gives better results. As a result, the NN is founded successful for the prediction of stress-strain curve of near b titanium alloy.