Purpose: To investigate the neural specificity of pericardium PC6, with the same meridian acupoint PC7 and a treatment‐irrelevant acupoint GB37 as separate controls.Materials and Methods: Functional magnetic resonance imaging (MRI) of the whole brain was performed in 36 healthy subjects receiving acupuncture at three acupoints, respectively: the study acupoint (PC6), and control acupoints (PC7 and GB37). A novel nonrepeated event‐related (NRER) design paradigm was applied to separately detect neural activities related to different stages of acupuncture (needling manipulation and post‐acupuncture rest epoch). Psychophysical responses (Deqi sensations) were also assessed.Results: Neuroimaging studies of PC6 presented extensive signal attenuations in the cerebrocerebellar and subcortical areas, whereas acupuncture at GB37 induced widespread signal potentiations. In addition, acupuncture at PC6, in comparison with stimulations at PC7 and GB37, selectively evoked neural responses of the insula, hypothalamus, and flocculonodular lobe of cerebellum (nodulus and uvula).Conclusion: These findings may provide preliminary evidence for specific involvements of the cerebellar‐hypothalamus and insula following acupuncture at PC6, which underlies the autonomic regulation of vestibular functions. The predominantly time‐prolonged deactivations in these areas may also serve the clinical efficacy of PC6 in producing a sedative or tranquilizing effect in antiemetic treatment. J. Magn. Reson. Imaging 2010;31:71–77.

Amnestic mild cognitive impairment (aMCI) is the transitional, heterogeneous continuum from normal elderly to Alzheimer's disease (AD). Previous studies have shown that brain functional activity in the default mode network (DMN) is impaired in aMCI patients with saliently cognitive and memory decline. However, the effective connectivity among the spatially isolated, but functionally related areas within the DMN in aMCI patients remains largely unknown. The present study examined dysfunctional connectivity of the DMN by combining an independent component analysis (ICA) approach with multivariate Granger causality analysis (mGCA) in 18 aMCI patients and 18 age-matched cognitively normal elderly. Results from mGCA showed decreased effective connectivity occurred among the middle temporal gyrus (MTG), hippocampus (HC) and fusiform gyrus (FG), as well as between the precuneus/posterior cingulate cortex (PreCN/PCC) and HC in patients with aMCI. Such an impaired connectivity was also correlated with patients’ cognitive performance of the auditory verbal learning. Moreover, enhanced effective connectivity within frontal cortex emerged, which may maintain memory functions after attenuated connections within DMN activity. These findings may elucidate the dysfunctional processes in brain networks of aMCI patients, highlighting the importance of connectivity changes in the pathophysiology of aMCI. (JINS, 2013, 19, 1–10)

Some patients after mild traumatic brain injury (mTBI) experience microstructural damages in the long-distance white matter (WM) connections, which disrupts the functional connectome of large-scale brain networks that support cognitive function. Patterns of WM structural damage following mTBI were well documented using diffusion tensor imaging (DTI). However, the functional organization of WM and its association with gray matter functional networks (GM-FNs) and its DTI metrics remain unknown. The present study adopted resting-state functional magnetic resonance imaging to explore WM functional properties in mTBI patients (108 acute patients, 48 chronic patients, 46 healthy controls [HCs]). Eleven large-scale WM functional networks (WM-FNs) were constructed by the k-means clustering algorithm of voxel-wise WM functional connectivity (FC). Compared with HCs, acute mTBI patients observed enhanced FC between inferior fronto-occipital fasciculus (IFOF) WM-FN and primary sensorimotor WM-FNs, and cortical primary sensorimotor GM-FNs. Further, acute mTBI patients showed increased DTI metrics (mean diffusivity, axial diffusivity, and radial diffusivity) in deep WM-FNs and higher-order cognitive WM-FNs. Moreover, mTBI patients demonstrated full recovery of FC and partial recovery of DTI metrics in the chronic stage. Additionally, enhanced FC between IFOF WM-FN and anterior cerebellar GM-FN was correlated with impaired information processing speed. Our findings provide novel evidence for functional and structural alteration of WM-FNs in mTBI patients. Importantly, the convergent damage of the IFOF network might imply its crucial role in our understanding of the pathophysiology mechanism of mTBI patients.

Speech reflects people’s mental state and using a microphone sensor is a potential method for human–computer interaction. Speech recognition using this sensor is conducive to the diagnosis of mental illnesses. The gender difference of speakers affects the process of speech emotion recognition based on specific acoustic features, resulting in the decline of emotion recognition accuracy. Therefore, we believe that the accuracy of speech emotion recognition can be effectively improved by selecting different features of speech for emotion recognition based on the speech representations of different genders. In this paper, we propose a speech emotion recognition method based on gender classification. First, we use MLP to classify the original speech by gender. Second, based on the different acoustic features of male and female speech, we analyze the influence weights of multiple speech emotion features in male and female speech, and establish the optimal feature sets for male and female emotion recognition, respectively. Finally, we train and test CNN and BiLSTM, respectively, by using the male and the female speech emotion feature sets. The results show that the proposed emotion recognition models have an advantage in terms of average recognition accuracy compared with gender-mixed recognition models.

Traumatic brain injuries (TBIs) are generally recognized to affect episodic memory. However, less is known regarding how external force altered the way functionally connected brain structures of the episodic memory system interact. To address this issue, we adopted an effective connectivity based analysis, namely, multivariate Granger causality approach, to explore causal interactions within the brain network of interest. Results presented that TBI induced increased bilateral and decreased ipsilateral effective connectivity in the episodic memory network in comparison with that of normal controls. Moreover, the left anterior superior temporal gyrus (aSTG, the concept forming hub), left hippocampus (the personal experience binding hub), and left parahippocampal gyrus (the contextual association hub) were no longer network hubs in TBI survivors, who compensated for hippocampal deficits by relying more on the right hippocampus (underlying perceptual memory) and the right medial frontal gyrus (MeFG) in the anterior prefrontal cortex (PFC). We postulated that the overrecruitment of the right anterior PFC caused dysfunction of the strategic component of episodic memory, which caused deteriorating episodic memory in mTBI survivors. Our findings also suggested that the pattern of brain network changes in TBI survivors presented similar functional consequences to normal aging.

We aimed to investigate the activity within and the connectivity between resting state networks (RSNs) in healthy subjects and patients with Alzheimer’s disease (AD) or mild cognitive impairment (MCI). Magnetic resonance imaging (MRI) and resting-state MRI were performed on patients diagnosed with AD (n=18) or MCI (n=16) and on healthy subjects (n=18) with matching demographic characteristics (age, sex, and education level). Independent component analysis and Granger causality analysis (GCA) were used during image postprocessing. We calculated ‘In + Out degree’ for each RSN. Then, we investigated the relationships between “In + Out degree” of each brain network and the cognitive behavioural data. RSNs were obtained using the optimal matching method. The core areas of the five RSNs were similar between the AD, MCI, and healthy control groups, but the activity within these five RSNs was significantly lower in the AD and MCI groups than in the healthy control group (P<0.01, false discovery rate corrected). The GCA results showed that the connectivity between the five RSNs, particularly the connectivity from the default mode network (DMN) to the other RSNs, was slightly lower in MCI patients and was significantly lower in AD patients than in healthy subjects. In contrast, increased connectivity was evident between the memory network and the executive control network in the AD and MCI patients. The “In + Out degree” of the DMN negatively correlated with the Montreal Cognitive Assessment score in AD patients (R=-0.43, P<0.05). In conclusion, the activity within RSNs and the connectivity between RSNs differed between AD patients, MCI patients, and normal individuals; these results provide an imaging reference for the diagnosis of AD and the measurement of disease progression and reveal insight into the pathogenesis of AD.