The aim of this study was to evaluate the capability of improved artificial neural networks (ANN) and additional novel training methods in distinguishing between benign and malignant breast lesions in contrast-enhanced magnetic resonance-mammography (MRM). A total of 604 histologically proven cases of contrast-enhanced lesions of the female breast at MRI were analyzed. Morphological, dynamic and clinical parameters were collected and stored in a database. The data set was divided into several groups using random or experimental methods [Training & Testing (T&T) algorithm] to train and test different ANNs. An additional novel computer program for input variable selection was applied. Sensitivity and specificity were calculated and compared with a statistical method and an expert radiologist. After optimization of the distribution of cases among the training and testing sets by the T & T algorithm and the reduction of input variables by the Input Selection procedure a highly sophisticated ANN achieved a sensitivity of 93.6% and a specificity of 91.9% in predicting malignancy of lesions within an independent prediction sample set. The best statistical method reached a sensitivity of 90.5% and a specificity of 68.9%. An expert radiologist performed better than the statistical method but worse than the ANN (sensitivity 92.1%, specificity 85.6%). Features extracted out of dynamic contrast-enhanced MRM and additional clinical data can be successfully analyzed by advanced ANNs. The quality of the resulting network strongly depends on the training methods, which are improved by the use of novel training tools. The best results of an improved ANN outperform expert radiologists.

The paper analyzes the fluid dynamic performance of a double inlet Gerotor pump by means of a multi-phase and multicomponent CFD approach. The numerical simulation includes the full 3D geometry of the pump as well as the real physics of the compressible hydraulic fluid and the rotating dynamic motion. The aeration and cavitation phenomena are included in the analysis adopting the Rayleight-Plesset equation and inertia controlled growth model for bubble formation. Cavitation and aeration phenomena are detected, especially when intake pressure is lower than atmospheric pressure. The influence of the fluid temperature variation on the component performance is also numerically predicted. The accuracy of a detailed modelling of the fluid properties variation with respect to the temperature and pressure is addressed and the effects on the numerical results is investigated. The rotational speeds of the internal and the external gears of the pump and the engagement between the teeth are addressed by means of an overset mesh approach. Constant leak height is considered between the gears and the case, while the overset mesh approach is adopted in order to accurately predict the leakage due to the teeth engagement. This numerical approach enables to investigate the dynamic performance of Gerotor gear pumps in terms of flow rate and pressure ripples and volumetric efficiency under standard and critical (actual) operating conditions. Good agreement between numerical and experimental results was found for specific operating conditions.

The paper investigates the oil flow through a multi plate clutch for a hydro-mechanical variable transmission under actual operating conditions. The analysis focuses on the numerical approach for the accurate prediction of the transient behavior of the lubrication in the gear region: the trade-off between prediction capabilities of the numerical model and computational effort is addressed. The numerical simulation includes the full 3D geometry of the clutch and the VOF multi-phase approach is used to calculate the oil distribution in the clutch region under different relative rotating velocities. Furthermore, the lubrication of the friction disks is calculated for different clutch actuation conditions, i.e. not-engaged and engaged positions. The influence of different geometrical features of the clutch lubricating circuit on the oil distribution is also determined. The results show the areas where poor lubrication occurs and extend the experiments where measurements are difficult to carry out. The simulation highlights the regions where high thermal stresses are observed during tests

The paper proposes a novel concept for axels dry braking system in off-road vehicles by implementing an oil recovery system in the friction plates chamber. The new system is able to remove the oil in the discs’ chamber when they are not engaged and to replenish it when the braking system is activated and the heat generated has to be dissipated. Thus, the energy losses due to the oil splashing will be significantly reduced with remarkable effects on the fuel consumption of the vehicle. Since experimental measurements are very difficult to carry out on a real system, a simplified geometry is designed and an ad-hoc test rig realized. Fast imaging techniques are used to capture the multiphase flow pattern within the friction plates chamber at different rotational speeds of the axel. The experimental results are used to validate a full 3D multi-phase CFD approach. A good agreement between the measurements and the calculations is found. The numerical modeling is therefore employed to predict the flow distribution in the real geometry and under actual operating conditions. A modular approach is adopted for the domain subdivision in order to represent accurately the three dimensional geometrical features, while the volume of fluid approach is used to model the multi-phase flow that characterizes the component. A conjugate heat transfer model is also adopted to predict the heat transferred from the discs to the working fluid and how the fluid is dissipating the heat within the component. By means of the numerical analysis the geometry of the real system is designed in order to improve the performance of the dry braking systems both in terms of energy saving and oil cooling.

No abstract