Complex extreme learning machine applications in terahertz pulsed signals feature sets

This paper presents a novel approach to the automatic classification of very large data sets composed of terahertz pulse transient signals, highlighting their potential use in biochemical, biomedical, pharmaceutical and security applications. Two different types of THz spectra are considered in the classification process. Firstly a binary classification study of poly-A and poly-C ribonucleic acid samples is performed. This is then contrasted with a difficult multi-class classification problem of spectra from six different powder samples that although have fairly indistinguishable features in the optical spectrum, they also possess a few discernable spectral features in the terahertz part of the spectrum. Classification is performed using a complex-valued extreme learning machine algorithm that takes into account features in both the amplitude as well as the phase of the recorded spectra. Classification speed and accuracy are contrasted with that achieved using a support vector machine classifier. The study systematically compares the classifier performance achieved after adopting different Gaussian kernels when separating amplitude and phase signatures. The two signatures are presented as feature vectors for both training and testing purposes. The study confirms the utility of complex-valued extreme learning machine algorithms for classification of the very large data sets generated with current terahertz imaging spectrometers. The classifier can take into consideration heterogeneous layers within an object as would be required within a tomographic setting and is sufficiently robust to detect patterns hidden inside noisy terahertz data sets. The proposed study opens up the opportunity for the establishment of complex-valued extreme learning machine algorithms as new chemometric tools that will assist the wider proliferation of terahertz sensing technology for chemical sensing, quality control, security screening and clinic diagnosis. Furthermore, the proposed algorithm should also be very useful in other applications requiring the classification of very large datasets.

The neurogenic potential of astrocytes is regulated by inflammatory signals

Although the adult brain contains neural stem cells (NSCs) that generate new neurons throughout life, these astrocyte-like populations are restricted to two discrete niches. Despite their terminally differentiated phenotype, adult parenchymal astrocytes can re-acquire NSC-like characteristics following injury, and as such, these 'reactive' astrocytes offer an alternative source of cells for central nervous system (CNS) repair following injury or disease. At present, the mechanisms that regulate the potential of different types of astrocytes are poorly understood. We used in vitro and ex vivo astrocytes to identify candidate pathways important for regulation of astrocyte potential. Using in vitro neural progenitor cell (NPC)-derived astrocytes, we found that exposure of more lineage-restricted astrocytes to either tumor necrosis factor alpha (TNF-α) (via nuclear factor-κB (NFκB)) or the bone morphogenetic protein (BMP) inhibitor, noggin, led to re-acquisition of NPC properties accompanied by transcriptomic and epigenetic changes consistent with a more neurogenic, NPC-like state. Comparative analyses of microarray data from in vitro-derived and ex vivo postnatal parenchymal astrocytes identified several common pathways and upstream regulators associated with inflammation (including transforming growth factor (TGF)-β1 and peroxisome proliferator-activated receptor gamma (PPARγ)) and cell cycle control (including TP53) as candidate regulators of astrocyte phenotype and potential. We propose that inflammatory signalling may control the normal, progressive restriction in potential of differentiating astrocytes as well as under reactive conditions and represent future targets for therapies to harness the latent neurogenic capacity of parenchymal astrocytes.