Depression Identification Using EEG Signals via a Hybrid of LSTM and Spiking Neural Networks.

Depression severity can be classified into distinct phases based on the Beck depression inventory (BDI) test scores, a subjective questionnaire. However, quantitative assessment of depression may be attained through the examination and categorization of electroencephalography (EEG) signals. Spiking neural networks (SNNs), as the third generation of neural networks, incorporate biologically realistic algorithms, making them ideal for mimicking internal brain activities while processing EEG signals. This study introduces a novel framework that for the first time, combines an SNN architecture and a long short-term memory (LSTM) structure to model the brain's underlying structures during different stages of depression and effectively classify individual depression levels using raw EEG signals. By employing a brain-inspired SNN model, our research provides fresh perspectives and advances knowledge of the neurological mechanisms underlying different levels of depression. The methodology employed in this study includes the utilization of the synaptic time dependent plasticity (STDP) learning rule within a 3-dimensional brain-template structured SNN model. Furthermore, it encompasses the tasks of classifying and predicting individual outcomes, visually representing the structural alterations in the brain linked to the anticipated outcomes, and offering interpretations of the findings. Notably, our method achieves exceptional accuracy in classification, with average rates of 98% and 96% for eyes-closed and eyes-open states, respectively. These results significantly outperform state-of-the-art deep learning methods.

Continuous Scoring of Depression From EEG Signals via a Hybrid of Convolutional Neural Networks.

Depression score is traditionally determined by taking the Beck depression inventory (BDI) test, which is a qualitative questionnaire. Quantitative scoring of depression has also been achieved by analyzing and classifying pre-recorded electroencephalography (EEG) signals. Here, we go one step further and apply raw EEG signals to a proposed hybrid convolutional and temporal-convolutional neural network (CNN-TCN) to continuously estimate the BDI score. In this research, the EEG signals of 119 individuals are captured by 64 scalp electrodes through successive eyes-closed and eyes-open intervals. Moreover, all the subjects take the BDI test and their scores are determined. The proposed CNN-TCN provides mean squared error (MSE) of 5.64±1.6 and mean absolute error (MAE) of 1.73±0.27 for eyes-open state and also provides MSE of 9.53±2.94 and MAE of 2.32±0.35 for the eyes-closed state, which significantly surpasses state-of-the-art deep network methods. In another approach, conventional EEG features are elicited from the EEG signals in successive frames and apply them to the proposed CNN-TCN in conjunction with known statistical regression methods. Our method provides MSE of 10.81±5.14 and MAE of 2.41±0.59 that statistically outperform the statistical regression methods. Moreover, the results with raw EEG are significantly better than those with EEG features.

Removal of eye blinking artifact from the electro-encephalogram, incorporating a new constrained blind source separation algorithm.

A robust constrained blind source separation (CBSS) algorithm has been developed as an effective means to remove ocular artifacts (OAs) from electro-encephalograms (EEGs). Currently, clinicians reject a data segment if the patient blinked or spoke during the observation interval. The rejected data segment could contain important information masked by the artifact. In the CBSS technique, a reference signal was exploited as a constraint. The constrained problem was then converted to an unconstrained problem by means of non-linear penalty functions weighted by the penalty terms. This led to the modification of the overall cost function, which was then minimised with the natural gradient algorithm. The effectiveness of the algorithm was also examined for the removal of other interfering signals such as electrocardiograms. The CBSS algorithm was tested with ten sets of data containing OAs. The proposed algorithm yielded, on average, a 19% performance improvement over Parra's BSS algorithm for removing OAs.

Localization of brain abnormal signal sources using blind source separation.

An effective and simple algorithm for localization of abnormal sources of the EEG signals within the brain has been developed here. In this method the signals are separated first, then the estimated independent components are lowpass filtered and normalized. In the next stage the correlation values between the estimated sources and the electrode signals are measured. On the other hand the sources with known locations are separated offline using narrowband bandpass filters. Finally, as the main contribution of the paper the mixing matrix is estimated using the information about the known sources and the estimated sources. The locations of the unknown sources are then measured with respect to the columns of the mixing matrix and the geometrical properties of the head and electrode locations.

Vegetative compatibility and pathogenecity of Verticillium dahliae kleb. isolates from olive in Iran.

Verticillium wilt caused by Verticillium dahliae is a serious problem of olive trees leading to significant reduction in yield. Verticillium wilt of olive trees was first recorded in Iran 1996 and confirm as due to Verticillium dahliae Kleb. 101 isolates of V. dahliae from olive trees at deferent locations in north provinces of Iran were assigned to vegetative compatibility groups (VCGS), using nitrate non-utilizing (Nit) mutants. A higher frequency of nit 1/nit 3 mutants (93%) was obtained compared with NitM (7%) with 10% of the isolates being assigned to VCG1 and 51% VCG4B and 19% VCG2A. 20% of isolates could not be classified in standard isolates. The pathogenecity of 15 randomly selected isolates (5 of each VCG) was tested on olive seedling (cv. Zard) and eggplant. The VCGs isolates were similarly aggressive on olive. However, VCG1 isolates were more aggressive on eggplant cv. Local than the VCG2A and VCG4B isolates as indicated by a higher colonization index. The pathogenecity tests of the pathogen on test plants (cotton cv. 'sahel', eggplant cv. 'local' and tomato cv. 'ps') show all isolates category in 2 pathogenecity groups defoliate and non-defoliate (with severe and mild subgroups). The morphology of V. dahliae isolates on C'zapeck's agar and water agar medium were different especially for microsclerotia appearance time in culture and their morphology.

Olive verticillium wilt or dieback of olive in Iran.

During 2000--03, different areas in Zanjan, Golestan and Khorasan provinces were surveyed for the presence of olive dieback. Olive branches, leaves and roots showing typical symptoms and soil around the roots were collected for further study. Samples were surface-sterilized with sodium hypochlorite or ethanol and then cultured on PDA and Czapek media. Soil samples were diluted in ethanol-agar for fungal isolation and purification. Morphological characteristics of the fungal mycelium particularly phialide and spores identified the causal agent to be the soil-borne pathogen, Verticillium dahliae. The disease was present in all olive growing regions but it was severe in temperate and relatively humid regions such as Gorgan. Infection index of the disease varied between 5 to 30% with an average of 11.89+/-1.12 among various orchards in this area. The newly established orchards showed more infection than the older ones. A significant difference in disease incidence and severity were observed among olive cultivars of Michen, Roughani, Zard and Koronakei. The latter cultivar had the least amount of infection. Strains of V. dahliae isolated from olive trees had different morphological and pathogenicity characteristics. These strains had different growth rates in response to the optimum temperature of 20 or 25 degrees C. The number of fungal propagules per gram of air-dried soil ranged from 2 to 32 with an average number of 13.42+/-0.50. Regarding the number of propagules of V. dahliae in the soil and susceptibility of cultivars in the newly established orchards, it seems necessary to take serious control measures to prevent disease spread.

Influence of inoculum density of Verticillium dahliae, temperature and relative humidity on epidemics of verticillium wilt of cotton in northern Iran.

During 1992--2003, frequency of Verticillium dahliae propagules, disease incidence and severity of verticillium wilt of cotton were determined in several cotton growing fields in Golestan province, northeastern Iran. Inoculum density varied among fields and different years ranging between 2-47 propagules/g of air-dried soil with an average of 18.96+/-0.73. In addition, the pattern of diseased plants varied with type of field and year. Simple regression analysis showed a linear relationship between inoculum density of V. dahliae at planting time on one hand, disease incidence and severity for all years on the other. The straight line model described the increase in disease intensity index over the accumulated physiological time from sowing. The number of days above 28 degrees C (T) and the area under relative humidity (RH) had significant effects on inoculum density in soil (MS) and final disease development (Y) and fitted the Y = 65.840 - 0.0034 RH + 0.57 MS - 1.7T model with R2 = 0.859 and significant F-function (p<0.0001).

Reconstruction of electroncephalogram brain maps by incorporating the blind source separation concept.

Electroencephalogram (EEG) brain maps provide useful and reliable neurodiagnostic information. Accurate reconstruction of the data requires an efficient separation of the electrode signals. Although autoregressive (AR) spectrum estimation highly refines the signals, it cannot remove the effect of adjacent electrode signals. This report describes an efficient blind signal separation (BSS) method. The algorithm identifies the coefficients of an adaptive FIR filter by minimization of a cost function in terms of the corresponding fourth-order cumulants. Applying this method, the quality of the results is far superior to the traditional methods.