E3GCAPS: Efficient EEG-Based Multi-Capsule Framework with Dynamic Attention for Cross-Subject Cognitive State Detection

Cognitive state detection using electroencephalogram(EEG)signals for various tasks has attracted significant research attention.However,it is difficult to further improve the performance of crosssubject cognitive state detection.Further,most of the existing deep learning models will degrade significantly when limited training samples are given,and the feature hierarchical relationships are ignored.To address the above challenges,we propose an efficient interpretation model based on multiple capsule networks for cross-subject EEG cognitive state detection,termed as Efficient EEG-based Multi-Capsule Framework(E3GCAPS).Specifically,we use a selfexpression module to capture the potential connections between samples,which is beneficial to alleviate the sensitivity of outliers that are caused by the individual differences of cross-subject EEG.In addition,considering the strong correlation between cognitive states and brain function connection mode,the dynamic subcapsule-based spatial attention mechanism is introduced to explore the spatial relationship of multi-channel 1D EEG data,in which multichannel 1D data greatly improving the training efficiency while preserving the model performance.The effectiveness of the E3GCAPS is validated on the Fatigue-Awake EEG Dataset(FAAD)and the SJTU Emotion EEG Dataset(SEED).Experimental results show E3GCAPS can achieve remarkable results on the EEG-based cross-subject cognitive state detection under different tasks.

Quantum Computational Techniques for Prediction of Cognitive State of Human Mind from EEG Signals

The utilization of quantum states for the representation of information and the advances in machine learning is considered as an efficient way of modeling the working of complex systems.The states of mind or judgment outcomes are highly complex phenomena that happen inside the human body.Decoding these states is significant for improving the quality of technology and providing an impetus to scientific research aimed at understanding the functioning of the human mind.One of the key advantages of quantum wave-functions over conventional classical models is the existence of configurable hidden variables,which provide more data density due to its exponential state-space growth.These hidden variables correspond to the amplitudes of each probable state of the system and allow for the modeling of various intricate aspects of measurable and observable physical quantities.This makes the quantum wave-functions powerful and felicitous to model cognitive states of the human mind,as it inherits the ability to efficiently couple the current context with past experiences temporally and spatially to approach an appropriate future cognitive state.This paper implements and compares some techniques like Variational Quantum Classifiers(VQC),quantum annealing classifiers,and hybrid quantum-classical neural networks,to harness the power of quantum computing for processing cognitive states of the mind by making use of EEG data.It also introduces a novel pipeline by logically combining some of the aforementioned techniques,to predict future cognitive responses.The preliminary results of these approaches are presented and are very encouraging with upto 61.53%validation accuracy.

Differences in cognitive profiles between traumatic brain injury and stroke： A comparison of the Montreal Cognitive Assessment and Mini-Mental State Examination

Purpose： To investigate the profiles of cognitive impairment through Montreal Cognitive Assessment （MoCA） and Mini-Mental State Examination （MMSE） in patients with chronic traumatic brain injury （TBI） or stroke and to evaluate the sensitivity of the two scales in patients with TBI. Methods： In this cohort study, a total of 230 patients were evaluated, including TBI group （n = 103） and stroke group （n - 127）. The cognitive functions of two groups were evaluated by designated specialists using MoCA （Beijing version） and MMSE （Chinese version）. Results： Compared with the patients with stroke, the patients with TBI received significantly lower score in orientation subtest and recall subtest in both tests. MoCA abnormal rates in the TBI group and stroke group were 94.17% and 86.61% respectively, while MMSE abnormal rates were 69.90% and 57.48%, respectively. In the TBI group, 87.10% patients with normal MMSE score had abnormal MoCA score and in the stroke group, about 70.37% patients with normal MMSE score had abnormal MoCA score. The diagnostic consistency of two scales in the TBI group and the stroke group were 72% and 69%, re.spectively. Conclusion： In our rehabilitation center, patients with TBI may have mare extensive and severe cognitive impairments than patients with stroke, prominently in orientation and recall domain. In screening post- TBI cognitive impairment, MoCA tends to be more sensitive than MIV[SE.

Brain activity in Parkinson's disease patients with mild cognitive impairment

Mild cognitive impairment （MCI） is common in patients with Parkinson＇s disease （PD）, yet the underlying neural mechanisms of this disease state remain unclear. We investigated alterations in the spontaneous brain activity of PD patients with MCI （PD-MCI） relative to cognitively normal PD patients （PD-CN） and healthy control （HC） subjects. In this work, 13 PD-MCI patients, 16 PD-CN patients, and 16 HC subjects completed resting state functional MRI. Spontaneous brain activity was measured by calculating amplitude of low frequency fluctuation （ALFF） values across the whole brain. Between-group differences and correlations between ALFF values and cognitive test scores were analyzed. ALFF values decreased in the right superior temporal gyrus and increased in the left middle temporal gyrus and left superior frontal gyms of PD-MCI patients compared with PD-CN patients. In the PD-MCI group, ALFF values in the left middle temporal gyrus were negatively correlated with Montreal Cognitive Assessment and vocabulary test scores, and the ALFF values in the left superior frontal gyms were negatively correlated with vocabulary test scores. Our study demonstrates that PD-MCI is associated with abnormal spontaneous brain activity in the temporal and frontal lobes. These findings inform the underlying neural mechanism of cognitive impairment in PD.