Leveraging network analysis to support experts in their analyses of subjects with MCI and AD

Optimal time-frequency localized wavelet filters for identification of Alzheimer's disease from EEG signals

Alzheimer's disease (AD) is a chronic disability that occurs due to the loss of neurons. The traditional methods to detect AD involve questionnaires and expensive neuro-imaging tests, which are time-consuming, subjective, and inconvenient to the target population. To overcome these limitations, Electroencephalogram (EEG) based methods have been developed to classify AD patients from normal controlled (NC) and mild cognitive impairment (MCI) subjects. Most of the EEG-based methods involved entropy-based feature extraction and discrete wavelet transform. However, the existing AD classification methods failed to provide promising classification accuracy. Here, we proposed a wavelet-machine learning (ML) framework to detect AD using a newly designed biorthogonal filter bank by optimization of frequency and time localization of triplet halfband filter banks (OTFL-THFB). The OTFL-THFB decomposes EEG signals into various EEG sub- bands. Hjorth Parameters (HP) and Higuchi's Fractal Dimension (HFD) have been investigated to extract features from each EEG subband. Subsequently, ML models are trained and tested using different features such as OTFL-THFB with HFD, OTFL-THFB with HP, and OTFL-THFB with HFD and HP used for detecting AD with 10-fold cross-validation. This method was applied to two publicly available datasets. Our model achieved an accuracy of 98.91 % for AD versus NC and 98.65 % for AD versus MCI versus NC using the least square support vector machine. Results indicate that this framework surpassed existing state-of-the-art techniques for classifying AD from NC.

STEADYNet: Spatiotemporal EEG analysis for dementia detection using convolutional neural network

Dementia is a neuro-degenerative disorder with a high death rate, mainly due to high human error, time, and cost of the current clinical diagnostic techniques. The existing dementia detection methods using hand-crafted electroencephalogram (EEG) signal features are unreliable. A convolution neural network using spatiotemporal EEG signals (STEADYNet) is presented to improve the dementia detection. The STEADYNet uses a multichannel temporal EEG signal as input. The network is grouped into feature extraction and classification components. The feature extraction comprises two convolution layers to generate complex features, a max-pooling layer to reduce the EEG signal's spatiotemporal redundancy, and a dropout layer to improve the network's generalization. The classification processes the feature extraction output nonlinearly using two fully-connected layers to generate salient features and a softmax layer to generate disease probabilities. Two publicly available multiclass datasets of dementia are used for evaluation. The STEADYNet outperforms existing automatic dementia detection methods with accuracies of 99.29 % , 99.65 % , and 92.25 % for Alzheimer's disease, mild cognitive impairment, and frontotemporal dementia, respectively. The STEADYNet has a low inference time and floating point operations, suitable for real-time applications. It may aid neurologists in efficient detection and treatment. A Python implementation of the STEADYNet is available at https://github.com/SandeepSangle12/STEADYNet.git.

LCADNet: a novel light CNN architecture for EEG-based Alzheimer disease detection

Alzheimer's disease (AD) is a progressive and incurable neurologi-cal disorder with a rising mortality rate, worsened by error-prone, time-intensive, and expensive clinical diagnosis methods. Automatic AD detection methods using hand-crafted Electroencephalogram (EEG) signal features lack accuracy and reliability. A lightweight convolution neural network for AD detection (LCADNet) is investigated to extract disease-specific features while reducing the detection time. The LCADNet uses two convolutional layers for extracting complex EEG features, two fully connected layers for selecting disease-specific features, and a softmax layer for predicting AD detection probability. A max-pooling layer interlaced between convolutional layers decreases the time-domain redundancy in the EEG signal. The efficiency of the LCADNet and four pre-trained models using transfer learning is compared using a publicly available AD detection dataset. The LCADNet shows the lowest computation complexity in terms of both the number of floating point operations and inference time and the highest classification performance across six measures. The generalization of the LCADNet is assessed by cross-testing it with two other publicly available AD detection datasets. It outperforms existing EEG-based AD detection methods with an accuracy of 98.50%. The LCADNet may be a valuable aid for neurologists and its Python implemen- tation can be found at github.com/SandeepSangle12/LCADNet.git.

Spatiotemporal Dynamics of Direct Carbon Emission and Policy Implication of Energy Transition for China’s Residential Consumption Sector by the Methods of Social Network Analysis and Geographically Weighted Regression

As China’s second largest energy-use sector, residential consumption has a great potential for carbon dioxide (CO2) reduction and energy saving or transition. Thus, here, using the methods of social network analysis (SNA) and geographically weighted regression (GWR), we investigated the spatiotemporal evolution characteristics of China’s residential CO2 emissions (RCEs) from direct energy use and proposed some policy suggestions for regional energy transition. (1) From 2000 to 2019, the total direct RCEs rose from 396.32 Mt to 1411.69 Mt; the consumption of electricity and coal were the primary sources. Controlling coal consumption and increasing the proportion of electricity generated from renewable energy should be the effective way of energy transition. (2) The spatial associations of direct RCEs show an obvious spatial network structure and the number of associations is increasing. Provinces with a higher level of economic development (Beijing, Shanghai, and Jiangsu) were at the center of the network and classified as the net beneficiary cluster in 2019. These provinces should be the priority areas of energy transition. (3) The net spillover cluster (Yunnan, Shanxi, Xinjiang, Gansu, Qinghai, Guizhou) is an important area to develop clean energy. People in this cluster should be encouraged to use more renewable energy. (4) GDP and per capita energy consumption had a significant positive influence on the growth of direct RCEs. Therefore, the national economy should grow healthily and sustainably to provide a favorable economic environment for energy transition. Meanwhile, residential consumption patterns should be greener to promote the use of clean energy.