Multi-level neural modelling : an application of object-oriented software engineering

Neu-Model, an ongoing project aimed at developing a neural simulation environment that is extremely computationally powerful and flexible, is described. It is shown that the use of good Software Engineering techniques in Neu-Model’s design and implementation is resulting in a high performance system that is powerful and flexible enough to allow rigorous exploration of brain function at a variety of conceptual levels.

Acoustic identification of twelve species of echolocating bat by discriminant function analysis and artificial neural networks

We recorded echolocation calls from 14 sympatric species of bat in Britain. Once digitised, one temporal and four spectral features were measured from each call. The frequency-time course of each call was approximated by fitting eight mathematical functions, and the goodness of fit, represented by the mean-squared error, was calculated. Measurements were taken using an automated process that extracted a single call from background noise and measured all variables without intervention. Two species of Rhinolophus were easily identified from call duration and spectral measurements. For the remaining 12 species, discriminant function analysis and multilayer back-propagation perceptrons were used to classify calls to species level. Analyses were carried out with and without the inclusion of curve-fitting data to evaluate its usefulness in distinguishing among species. Discriminant function analysis achieved an overall correct classification rate of 79% with curve-fitting data included, while an artificial neural network achieved 87%. The removal of curve-fitting data improved the performance of the discriminant function analysis by 2 %, while the performance of a perceptron decreased by 2 %. However, an increase in correct identification rates when curve-fitting information was included was not found for all species. The use of a hierarchical classification system, whereby calls were first classified to genus level and then to species level, had little effect on correct classification rates by discriminant function analysis but did improve rates achieved by perceptrons. This is the first published study to use artificial neural networks to classify the echolocation calls of bats to species level. Our findings are discussed in terms of recent advances in recording and analysis technologies, and are related to factors causing convergence and divergence of echolocation call design in bats.

Fine-grained plant classification using convolutional neural networks for feature extraction

We present an overview of the QUT plant classification system submitted to LifeCLEF 2014. This system uses generic features extracted from a convolutional neural network previously used to perform general object classification. We examine the effectiveness of these features to perform plant classification when used in combination with an extremely randomised forest. Using this system, with minimal tuning, we obtained relatively good results with a score of 0:249 on the test set of LifeCLEF 2014.

Whole-brain imaging with single-cell resolution using chemical cocktails and computational analysis

Systems-level identification and analysis of cellular circuits in the brain will require the development of whole-brain imaging with single-cell resolution. To this end, we performed comprehensive chemical screening to develop a whole-brain clearing and imaging method, termed CUBIC (clear, unobstructed brain imaging cocktails and computational analysis). CUBIC is a simple and efficient method involving the immersion of brain samples in chemical mixtures containing aminoalcohols, which enables rapid whole-brain imaging with single-photon excitation microscopy. CUBIC is applicable to multicolor imaging of fluorescent proteins or immunostained samples in adult brains and is scalable from a primate brain to subcellular structures. We also developed a whole-brain cell-nuclear counterstaining protocol and a computational image analysis pipeline that, together with CUBIC reagents, enable the visualization and quantification of neural activities induced by environmental stimulation. CUBIC enables time-course expression profiling of whole adult brains with single-cell resolution.

Prediction of GMA welding characteristic parameter by artificial neural network system

Nowadays, demand for automated Gas metal arc welding (GMAW) is growing and consequently need for intelligent systems is increased to ensure the accuracy of the procedure. To date, welding pool geometry has been the most used factor in quality assessment of intelligent welding systems. But, it has recently been found that Mahalanobis Distance (MD) not only can be used for this purpose but also is more efficient. In the present paper, Artificial Neural Networks (ANN) has been used for prediction of MD parameter. However, advantages and disadvantages of other methods have been discussed. The Levenberg–Marquardt algorithm was found to be the most effective algorithm for GMAW process. It is known that the number of neurons plays an important role in optimal network design. In this work, using trial and error method, it has been found that 30 is the optimal number of neurons. The model has been investigated with different number of layers in Multilayer Perceptron (MLP) architecture and has been shown that for the aim of this work the optimal result is obtained when using MLP with one layer. Robustness of the system has been evaluated by adding noise into the input data and studying the effect of the noise in prediction capability of the network. The experiments for this study were conducted in an automated GMAW setup that was integrated with data acquisition system and prepared in a laboratory for welding of steel plate with 12 mm in thickness. The accuracy of the network was evaluated by Root Mean Squared (RMS) error between the measured and the estimated values. The low error value (about 0.008) reflects the good accuracy of the model. Also the comparison of the predicted results by ANN and the test data set showed very good agreement that reveals the predictive power of the model. Therefore, the ANN model offered in here for GMA welding process can be used effectively for prediction goals.

The neural correlates of justified and unjustified killing: An fMRI study

Despite moral prohibitions on hurting other humans, some social contexts allow for harmful actions such as the killing of others. One example is warfare, where killing enemy soldiers is seen as morally justified. Yet, the neural underpinnings distinguishing between justified and unjustified killing are largely unknown. To improve understanding of the neural processes involved in justified and unjustified killing, participants had to imagine being the perpetrator whilst watching “first-person perspective” animated videos where they shot enemy soldiers (‘justified violence’) and innocent civilians (‘unjustified violence’). When participants imagined themselves shooting civilians compared to soldiers, greater activation was found in the lateral orbitofrontal cortex (OFC). Regression analysis revealed that the more guilt participants felt about shooting civilians, the greater the response in the lateral OFC. Effective connectivity analyses further revealed an increased coupling between lateral OFC and the tempoparietal junction (TPJ) when shooting civilians. The results show that the neural mechanisms typically implicated with harming others, such as the OFC, become less active when the violence against a particular group is seen as justified. This study therefore provides unique insight into how normal individuals can become aggressors in specific situations.

Autonomous helicopter hover using an artificial neural network

Details the developments to date of an unmanned air vehicle (UAV) based on a standard size 60 model helicopter. The design goal is to have the helicopter achieve stable hover with the aid of an INS and stereo vision. The focus of the paper is on the development of an artificial neural network (ANN) that makes use of only the INS data to generate hover commands, which are used to directly manipulate the flight servos. Current results show that networks incorporating some form of recurrency (state history) offer little advantage over those without. At this stage, the ANN has partially maintained periods of hover even with misaligned sensors.

Stimulation of osteogenesis and angiogenesis of hBMSCs by delivering Si ions and functional drug from mesoporous silica nanospheres

Multifunctional bioactive materials with the ability to stimulate osteogenesis and angiogenesis of stem cells play an important role in the regeneration of bone defects. However, how to develop such biomaterials remains a significant challenge. In this study, we prepared mesoporous silica nanospheres (MSNs) with uniform sphere size (∼90 nm) and mesopores (∼2.7 nm), which could release silicon ions (Si) to stimulate the osteogenic differentiation of human bone marrow stromal cells (hBMSCs) via activating their ALP activity, bone-related gene and protein (OCN, RUNX2 and OPN) expression. Hypoxia-inducing therapeutic drug, dimethyloxaloylglycine (DMOG), was effectively loaded in the mesopores of MSNs (D-MSNs). The sustained release of DMOG from D-MSNs could stabilize HIF-1α and further stimulated the angiogenic differentiation of hBMSCs as indicated by the enhanced VEGF secretion and protein expression. Our study revealed that D-MSNs could combine the stimulatory effect on both osteogenic and angiogenic activity of hBMSCs. The potential mechanism of D-MSN-stimulated osteogenesis and angiogenesis was further elucidated by the supplementation of cell culture medium with pure Si ions and DMOG. Considering the easy handling characteristics of nanospheres, the prepared D-MSNs may be applied in the forms of injectable spheres for minimally invasive surgery, or MSNs/polymer composite scaffolds for bone defect repair. The concept of delivering both stimulatory ions and functional drugs may offer a new strategy to construct a multifunctional biomaterial system for bone tissue regeneration.

Medical semantic similarity with a neural language model

Advances in neural network language models have demonstrated that these models can effectively learn representations of words meaning. In this paper, we explore a variation of neural language models that can learn on concepts taken from structured ontologies and extracted from free-text, rather than directly from terms in free-text. This model is employed for the task of measuring semantic similarity between medical concepts, a task that is central to a number of techniques in medical informatics and information retrieval. The model is built with two medical corpora (journal abstracts and patient records) and empirically validated on two ground-truth datasets of human-judged concept pairs assessed by medical professionals. Empirically, our approach correlates closely with expert human assessors ($\approx$ 0.9) and outperforms a number of state-of-the-art benchmarks for medical semantic similarity. The demonstrated superiority of this model for providing an effective semantic similarity measure is promising in that this may translate into effectiveness gains for techniques in medical information retrieval and medical informatics (e.g., query expansion and literature-based discovery).

The ghost in the machine: Legal challenges of neural interface devices

Neural interface devices and the melding of mind and machine, challenge the law in determining where civil liability for injury, damage or loss should lie. The ability of the human mind to instruct and control these devices means that in a negligence action against a person with a neural interface device, determining the standard of care owed by him or her will be of paramount importance. This article considers some of the factors that may influence the court’s determination of the appropriate standard of care to be applied in this situation, leading to the conclusion that a new standard of care might evolve.

Neural correlates of semantic priming for ambiguous words: An event-related fMRI study

We investigated the neural correlates of semantic priming by using event-related fMRI to record blood oxygen level dependent (BOLD) responses while participants performed speeded lexical decisions (word/nonword) on visually presented related versus unrelated prime-target pairs. A long stimulus onset asynchrony of 1000 ms was employed, which allowed for increased controlled processing and selective frequency-based ambiguity priming. Conditions included an ambiguous word prime (e.g. bank) and a target related to its dominant (e.g. money) or subordinate meaning (e.g. river). Compared to an unrelated condition, primed dominant targets were associated with increased activity in the LIFG, the right anterior cingulate and superior temporal gyrus, suggesting postlexical semantic integrative mechanisms, while increased right supramarginal activity for the unrelated condition was consistent with expectancy based priming. Subordinate targets were not primed and were associated with reduced activity primarily in occipitotemporal regions associated with word recognition, which may be consistent with frequency-based meaning suppression. These findings provide new insights into the neural substrates of semantic priming and the functional-anatomic correlates of lexical ambiguity suppression mechanisms.

Selectivity of human retinotopic visual cortex to S-cone-opponent, L/M-cone-opponent and achromatic stimulation

Our aim was to make a quantitative comparison of the response of the different visual cortical areas to selective stimulation of the two different cone-opponent pathways [long- and medium-wavelength (L/M)- and short-wavelength (S)-cone-opponent] and the achromatic pathway under equivalent conditions. The appropriate stimulus-contrast metric for the comparison of colour and achromatic sensitivity is unknown, however, and so a secondary aim was to investigate whether equivalent fMRI responses of each cortical area are predicted by stimulus contrast matched in multiples of detection threshold that approximately equates for visibility, or direct (cone) contrast matches in which psychophysical sensitivity is uncorrected. We found that the fMRI response across the two colour and achromatic pathways is not well predicted by threshold-scaled stimuli (perceptual visibility) but is better predicted by cone contrast, particularly for area V1. Our results show that the early visual areas (V1, V2, V3, VP and hV4) all have robust responses to colour. No area showed an overall colour preference, however, until anterior to V4 where we found a ventral occipital region that has a significant preference for chromatic stimuli, indicating a functional distinction from earlier areas. We found that all of these areas have a surprisingly strong response to S-cone stimuli, at least as great as the L/M response, suggesting a relative enhancement of the S-cone cortical signal. We also identified two areas (V3A and hMT+) with a significant preference for achromatic over chromatic stimuli, indicating a functional grouping into a dorsal pathway with a strong magnocellular input.

Neural activity associated with semantic versus phonological anomia treatments in aphasia

Naming impairments in aphasia are typically targeted using semantic and/or phonologically based tasks. However, it is not known whether these treatments have different neural mechanisms. Eight participants with aphasia received twelve treatment sessions using an alternating treatment design, with fMRI scans pre- and post-treatment. Half the sessions employed Phonological Components Analysis (PCA), and half the sessions employed Semantic Feature Analysis (SFA). Pre-treatment activity in the left caudate correlated with greater immediate treatment success for items treated with SFA, whereas recruitment of the left supramarginal gyrus and right precuneus post-treatment correlated with greater immediate treatment success for items treated with PCA. The results support previous studies that have found greater treatment outcome to be associated with activity in predominantly left hemisphere regions, and suggest that different mechanisms may be engaged dependent on the type of treatment employed.

Improving deep convultional neural networks with unsupervised feature learning

The latest generation of Deep Convolutional Neural Networks (DCNN) have dramatically advanced challenging computer vision tasks, especially in object detection and object classification, achieving state-of-the-art performance in several computer vision tasks including text recognition, sign recognition, face recognition and scene understanding. The depth of these supervised networks has enabled learning deeper and hierarchical representation of features. In parallel, unsupervised deep learning such as Convolutional Deep Belief Network (CDBN) has also achieved state-of-the-art in many computer vision tasks. However, there is very limited research on jointly exploiting the strength of these two approaches. In this paper, we investigate the learning capability of both methods. We compare the output of individual layers and show that many learnt filters and outputs of the corresponding level layer are almost similar for both approaches. Stacking the DCNN on top of unsupervised layers or replacing layers in the DCNN with the corresponding learnt layers in the CDBN can improve the recognition/classification accuracy and training computational expense. We demonstrate the validity of the proposal on ImageNet dataset.