Women with hypoactive sexual desire disorder compared to normal females: a functional magnetic resonance imaging study.

Lack of sexual interest is the most common sexual complaint among women. However, factors affecting sexual desire in women have rarely been studied. While the role of the brain in integrating the sensory, attentional, motivational, and motor aspects of sexual response is commonly acknowledged as important, little is known about specific patterns of brain activation and sexual interest or response, particularly among women. We compared 20 females with no history of sexual dysfunction (NHSD) to 16 women with hypoactive sexual desire disorder (HSDD) in a functional magnetic resonance imaging (fMRI) study that included assessment of subjective sexual arousal, peripheral sexual response using a vaginal photoplethysmograph (VPP), as well as brain activation across three time points. Video stimuli included erotic, sports, and relaxing segments. Subjective arousal to erotic stimuli was significantly greater in NHSD participants compared with HSDD. In the erotic-sports contrast, NHSD women showed significantly greater activation in the bilateral entorhinal cortex than HSDD women. In the same contrast, HSDD females demonstrated higher activation than NHSD females in the medial frontal gyrus (Brodmann area (BA) 10), right inferior frontal gyrus (BA 47) and bilateral putamen. There were no between group differences in VPP-correlated brain activation and peripheral sexual response was not significantly associated with either subjective sexual response or brain activation patterns. Findings were consistent across the three experimental sessions. The results suggest differences between women with NHSD and HSDD in encoding arousing stimuli, retrieval of past erotic experiences, or both. The findings of greater activation in BA 10 and BA 47 among women with HSDD suggest that this group allocated significantly more attention to monitoring and/or evaluating their responses than NHSD participants, which may interfere with normal sexual response.

New Experimental Equipment Recreating Geo-Reservoir Conditions in Large, Fractured, Porous Samples to Investigate Coupled Thermal, Hydraulic and Polyaxial Stress Processes.

Use of the subsurface for energy resources (enhanced geothermal systems, conventional and unconventional hydrocarbons), or for storage of waste (CO2, radioactive), requires the prediction of how fluids and the fractured porous rock mass interact. The GREAT cell (Geo-Reservoir Experimental Analogue Technology) is designed to recreate subsurface conditions in the laboratory to a depth of 3.5 km on 200 mm diameter rock samples containing fracture networks, thereby enabling these predictions to be validated. The cell represents an important new development in experimental technology, uniquely creating a truly polyaxial rotatable stress field, facilitating fluid flow through samples, and employing state of the art fibre optic strain sensing, capable of thousands of detailed measurements per hour. The cell's mechanical and hydraulic operation is demonstrated by applying multiple continuous orientations of principal stress to a homogeneous benchmark sample, and to a fractured sample with a dipole borehole fluid fracture flow experiment, with backpressure. Sample strain for multiple stress orientations is compared to numerical simulations validating the operation of the cell. Fracture permeability as a function of the direction and magnitude of the stress field is presented. Such experiments were not possible to date using current state of the art geotechnical equipment.

A hybrid linear/nonlinear training algorithm for feedforward neural networks.

This paper presents a new hybrid optimization strategy for training feedforward neural networks. The algorithm combines gradient-based optimization of nonlinear weights with singular value decomposition (SVD) computation of linear weights in one integrated routine. It is described for the multilayer perceptron (MLP) and radial basis function (RBF) networks and then extended to the local model network (LMN), a new feedforward structure in which a global nonlinear model is constructed from a set of locally valid submodels. Simulation results are presented demonstrating the superiority of the new hybrid training scheme compared to second-order gradient methods. It is particularly effective for the LMN architecture where the linear to nonlinear parameter ratio is large.