A non-contact continuous blood pressure measurement method based on video stream / 生物医学工程学杂志

Hypertension is the primary disease that endangers human health. A convenient and accurate blood pressure measurement method can help to prevent the hypertension. This paper proposed a continuous blood pressure measurement method based on facial video signal. Firstly, color distortion filtering and independent component analysis were used to extract the video pulse wave of the region of interest in the facial video signal, and the multi-dimensional feature extraction of the pulse wave was preformed based on the time-frequency domain and physiological principles; Secondly, an integrated feature selection method was designed to extract the universal optimal feature subset; After that, we compared the single person blood pressure measurement models established by Elman neural network based on particle swarm optimization, support vector machine (SVM) and deep belief network; Finally, we used SVM algorithm to build a general blood pressure prediction model, which was compared and evaluated with the real blood pressure value. The experimental results showed that the blood pressure measurement results based on facial video were in good agreement with the standard blood pressure values. Comparing the estimated blood pressure from the video with standard blood pressure value, the mean absolute error (MAE) of systolic blood pressure was 4.9 mm Hg with a standard deviation (STD) of 5.9 mm Hg, and the MAE of diastolic blood pressure was 4.6 mm Hg with a STD of 5.0 mm Hg, which met the AAMI standards. The non-contact blood pressure measurement method based on video stream proposed in this paper can be used for blood pressure measurement.

Fetal electrocardiogram signal extraction and analysis method combining fast independent component analysis algorithm and convolutional neural network / 生物医学工程学杂志

Fetal electrocardiogram (ECG) signals provide important clinical information for early diagnosis and intervention of fetal abnormalities. In this paper, we propose a new method for fetal ECG signal extraction and analysis. Firstly, an improved fast independent component analysis method and singular value decomposition algorithm are combined to extract high-quality fetal ECG signals and solve the waveform missing problem. Secondly, a novel convolutional neural network model is applied to identify the QRS complex waves of fetal ECG signals and effectively solve the waveform overlap problem. Finally, high quality extraction of fetal ECG signals and intelligent recognition of fetal QRS complex waves are achieved. The method proposed in this paper was validated with the data from the PhysioNet computing in cardiology challenge 2013 database of the Complex Physiological Signals Research Resource Network. The results show that the average sensitivity and positive prediction values of the extraction algorithm are 98.21% and 99.52%, respectively, and the average sensitivity and positive prediction values of the QRS complex waves recognition algorithm are 94.14% and 95.80%, respectively, which are better than those of other research results. In conclusion, the algorithm and model proposed in this paper have some practical significance and may provide a theoretical basis for clinical medical decision making in the future.

The correlation between brain functional network connectivity and inflammatory cytokines in patients with bipolar disorder II depressive episodes / 中华行为医学与脑科学杂志

Objective:To investigate the functional connectivity of default mode network (DMN) and limbic system, the expression level of inflammatory cytokine and their correlation in bipolar disorder type Ⅱ(BDⅡ) patients with depressive episodes.Methods:Thirty-three BD Ⅱ patients with depressive episodes and forty-six healthy controls were recruited to complete the resting-state functional magnetic resonance imaging (rs-fMRI). After image preprocessing, the DMN and limbic system were extracted from the image data by independent component analysis (ICA), so as to compare the differences of functional connectivity of resting brain network between the patients and the controls.Serum levels of inflammatory cytokines interleukin-6 (IL-6), interleukin-8(IL-8), interleukin-10(IL-10), tumor necrosis factor-α (TNF-α), and C-C motif chemokine ligand 4 (CCL4) in patients and healthy controls were detected.The correlation between functional connectivity of different brain regions and inflammatory cytokines was analyzed.SPSS 17.0 software was used for data statistical analysis.The two samples were compared using t-test or Mann-Whitney U-test, and Spearman was used for correlation testing. Results:In BDⅡ patients, the functional connectivity of the right medial prefrontal cortex(cluster-size=7 voxel, cluster-level PGRF<0.05, MNI: x=6, y=54, z=9, t=-3.765) and the left superior frontal gyrus(cluster-size=10 voxel, cluster-level PGRF<0.05, MNI: x=-21, y=54, z=15, t=-4.139) in DMN decreased, while the left cerebellum Ⅳ and Ⅴ lobules of limbic system (cluster-size=21 voxel, cluster-level PGRF<0.05, MNI: x=-15, y=-24, z=-30, t=4.468) and cerebellar tonsil of left cerebellum posterior lobe(cluster-size=8 voxel, cluster-level PGRF<0.05, MNI: x=-15, y=-51, z=-45, t=4.138) in the limbic system increased.Compared with the healthy controls, the serum levels of IL-10(7.39 (6.33, 9.32) pg/mL vs 6.54 (5.84, 7.39) pg/mL, Z=-2.937, P=0.003)and CCL4 (39.31 (25.77, 68.70) pg/mL vs 31.30 (20.32, 40.89) pg/mL, Z=-2.209, P=0.027) were higher in BDⅡ patients.The functional connectivity of the left cerebellum Ⅳ and Ⅴ lobules was positively correlated with the serum levels of IL-10 ( r=0.432, P=0.031) and that of the cerebellar tonsil of left cerebellum posterior lobe was positively correlated with the serum levels of IL-10 ( r=0.429, P=0.032) and CCL4 ( r=0.402, P=0.046). Conclusion:The functional connectivity of DMN and limbic system in BDⅡ patients with depressive episode is abnormal in resting-state fMRI.The expression level of inflammatory cytokines in patients' serum increases, and has correlation with the functional connection of limbic system.

Functional connectivity analysis of sensory motor network in patients with bipolar disorder type Ⅰ / 中华行为医学与脑科学杂志

Objective:To analyze the functional connectivity (FC) characteristics of sensory motor network (SMN) in patients with bipolar disorder type Ⅰ (BD-Ⅰ) by independent component analysis (ICA), and explore the correlation between abnormal SMN and clinical symptoms.Methods:Eighteen patients with BD-Ⅰ (BD-Ⅰ group) and 20 matched normal controls (HC group) were included.Both groups received resting state fMRI (rs-fMRI) scanning.Based on ICA-fMRI data, one-sample t-test and two-sample t-test were used to analyze the components of SMN and to explore abnormal brain regions between the two groups.Functional network analysis (FNC) was also used to explore the functional connectivity between SMN and other brain networks.Pearson correlation analysis were conducted by SPSS 17.0 to measure the potential associations between intra-and inter-network functional connectivity and age, education, score of Bech-Rafaelsen mania rating scale (BRMS), score of positive and negative syndrome scale (PANSS) and other indicators. Results:In BD-Ⅰ group, the functional connection in the right paracentral lobule (MIN: x=8, y=-32, z=68, t=4.86, P<0.001) and the right postcentral gyrus (MIN: x=41, y=-26, z=53, t=3.33, P<0.001) in SMN were higher than those in HC group.Compared with HC group, the connectivity value in patients with BD-Ⅰ increased between SMN-DAN (0.247±0.073, -0.078±0.080, t=-2.974, P<0.01, FDR adjusted), while the connectivity value decreased between SMN-DMN(-0.037±0.054, 0.272±0.067, t=3.520, P<0.01, FDR adjusted) and between SMN-rFPN(-0.034±0.055, 0.231±0.070, t=2.939, P<0.01, FDR adjusted). Conclusion:The sensorimotor network of patients with BD-Ⅰ has abnormal functional connections within and between networks, and FC values in some networks are positively correlated with manic symptoms, which may be part of the neural mechanisms of patients with BD-Ⅰ.

Study on large-scale resting-state functional networks in first-episode untreated adult patients with major depressive disorder / 中华行为医学与脑科学杂志

Objective:To investigate the value of large-scale brain network research based on independent component analysis (ICA) in discovering the changes of intrinsic functional connections within and between resting-state networks (RSNs) in first-episode untreated adult patients with major depressive disorder (MDD).Methods:From January 2019 to June 2021, twenty-three patients with MDD (MDD group) and 30 healthy volunteers (HC group) matched with gender, age and years of education were selected. All participants underwent resting-state brain function imaging (rs-fMRI), and the MDD group completed the 17-item Hamilton depression scale(HAMD-17). The independent component analysis (ICA) method was used to analyze rs-fMRI data, and meaningful RSNs were obtained. SPM12 and Gift softwares were used to compare the strength of intrinsic functional connection within and between the RSNs of the MDD group and HC group, and the Pearson correlation analysis was conducted by IBM SPSS statistics 25.0 to evaluate the correlation between the functional connection strength and HAMD-17 scores in MDD group.Results:Compared with the HC group, intrinsic functional connection strength of medial prefrontal cortex (mPFC) (MNI: x, y, z=-6, 54, 25)in MDD group was significantly enhanced, while the intrinsic functional connection strength of the left angular gyrus (AG) (MNI: x, y, z=-48, -66, 21), the left precuneus (PCu) (MNI: x, y, z=-6, -63, 33), the left dorsolateral prefrontal cortex (dlPFC) (MNI: x, y, z=-36, 12, 51)and the right anterior insula (AI) (MNI: x, y, z=36, 21, 0)were significantly weakened. Compared with the HC group, functional connection strength between posterior default mode net work(pDMN) and anterior default mode network(aDMN) in MDD group was significantly weakened ( t=-2.206, P=0.032), and function connection strength between pDMN and left frontal parietal network(lFPN) was significantly strengthened ( t=2.318, P=0.025). In MDD group, intrinsic functional connection strength of mPFC and the functional connection strength of pDMN-lFPN were positively correlated with the HAMD-17 score ( r=0.524, P=0.010; r=0.441, P=0.035). Conclusion:Large-scale brain network study based on the ICA can find abnormal functional connections within and between RSNs in first-episode untreated adult patients with MDD, and provide objective imaging markers for the clinical diagnosis and treatment of MDD.

Automatic removal algorithm of electrooculographic artifacts in non-invasive brain-computer interface based on independent component analysis / 生物医学工程学杂志

The non-invasive brain-computer interface (BCI) has gradually become a hot spot of current research, and it has been applied in many fields such as mental disorder detection and physiological monitoring. However, the electroencephalography (EEG) signals required by the non-invasive BCI can be easily contaminated by electrooculographic (EOG) artifacts, which seriously affects the analysis of EEG signals. Therefore, this paper proposed an improved independent component analysis method combined with a frequency filter, which automatically recognizes artifact components based on the correlation coefficient and kurtosis dual threshold. In this method, the frequency difference between EOG and EEG was used to remove the EOG information in the artifact component through frequency filter, so as to retain more EEG information. The experimental results on the public datasets and our laboratory data showed that the method in this paper could effectively improve the effect of EOG artifact removal and improve the loss of EEG information, which is helpful for the promotion of non-invasive BCI.

Decision Tree Based Salp Swarm Optimization for Multi Medical Data Classification with Feature Reduction Technique

Abstract The ambitious task in the domain of medical informatics is medical data classification. From medical datasets, intention to ameliorate human burden with the medical data classification entails to taking in classification designs. The medical data classification is the major focus of this paper, where a Decision Tree based Salp Swarm Optimization (DT-SWO) algorithm is proposed. After pre-processingthe hybrid feature selection method selects the medical data features. The high dimensional features are reduced by Discriminant Independent Component Analysis (DICA) and DT-SWO is to classify the most relevant class of medical data. The details of four datasets namely Leukemia, Diffuse Larger B-cell Lymphomas (DLBCL), Lung cancer and Colon relating to four diseases for heart, liver, cancer and lungs are collected from the UCI machine learning repository. Ultimately, the experimental outcomes demonstrated that the proposed DT-SWO algorithm is suitable for medical data classification than other algorithms.

Independent Component Analysis and Graph Theoretical Analysis in Patients with Narcolepsy / 神经科学通报·英文版

The present study was aimed to evaluate resting-state functional connectivity and topological properties of brain networks in narcolepsy patients compared with healthy controls. Resting-state fMRI was performed in 26 adult narcolepsy patients and 30 matched healthy controls. MRI data were first analyzed by group independent component analysis, then a graph theoretical method was applied to evaluate the topological properties in the whole brain. Small-world network parameters and nodal topological properties were measured. Altered topological properties in brain areas between groups were selected as region-of-interest seeds, then the functional connectivity among these seeds was compared between groups. Partial correlation analysis was performed to evaluate the relationship between the severity of sleepiness and functional connectivity or topological properties in the narcolepsy patients. Twenty-one independent components out of 48 were obtained. Compared with healthy controls, the narcolepsy patients exhibited significantly decreased functional connectivity within the executive and salience networks, along with increased functional connectivity in the bilateral frontal lobes within the executive network. There were no differences in small-world network properties between patients and controls. The altered brain areas in nodal topological properties between groups were mainly in the inferior frontal cortex, basal ganglia, anterior cingulate, sensory cortex, supplementary motor cortex, and visual cortex. In the partial correlation analysis, nodal topological properties in the putamen, anterior cingulate, and sensory cortex as well as functional connectivity between these regions were correlated with the severity of sleepiness (sleep latency, REM sleep latency, and Epworth sleepiness score) among narcolepsy patients. Altered connectivity within the executive and salience networks was found in narcolepsy patients. Functional connection changes between the left frontal cortex and left caudate nucleus may be one of the parameters describing the severity of narcolepsy. Changes in the nodal topological properties in the left putamen and left posterior cingulate, changes in functional connectivity between the left supplementary motor area and right occipital as well as in functional connectivity between the left anterior cingulate gyrus and bilateral postcentral gyrus can be considered as a specific indicator for evaluating the severity of narcolepsy.

Multi?modal evaluation of Alzheimer disease by using joint independent component analysis of functional MRI diffusion tensor imaging / 中华放射学杂志

Objective We utilized a joint independent component analysis (Joint ICA), a novel method that combined rs?fMRI and DTI information, to describe comprehensive characteristics of brain functional activities and microstructural changes in the continuum of AD. Methods We employed a Joint ICA to calculate ALFF maps of fMRI data and FA maps of DTI data and fuse them in healthy controls (n=68), SCD (n=35), amnesic MCI (n=47) and AD (n=31). Besides, we applied one way ANOVA to detect the significant differences of joint components among groups, while controlling the age, gender, education, head motion, volumes of gray matter, white matter and CSF. Partial correlation analysis was used to test the relationships between joint ICs and cognitive measures. Results The results showed that there was no inner?group difference in HC and SCD groups (F=14.16, P<0.05). Compared to HC, SCD and AD groups, the ALFF component of aMCI group showed higher values in the bilateral cerebellum, bilateral precuneus, bilateral angular gyrus, bilateral frontal gyrus, bilateral temporal areas, thalamus and left insula. And in these regions, the ALFF of AD group was lower than HC. For the FA component map, same differences were found in the corpus callosum and limbic system. Furthermore, positive partial correlation between the IC weights and Mini?Mental State Examination (MMSE) scores was also found (r=0.29, P<0.01). Conclusions Multi?modal evaluation of AD has been implemented by using Joint ICA analysis of fMRI?DTI, which would contribute to early prediction, diagnosis, and even effective intervention in AD. These findings could help to explain the underlying mechanism of the disease progression.

Resting-state Neural Networks and Granger Causal Connectivity for Patients with Cognitive Impairment Associated with Leukoaraiosis / 中国康复理论与实践

Objective:To compare the difference in resting state networks among leukoaraiosis (LA) patients with or without mild cognitive impairment, and healthy controls, as well as the functional connectivity under Granger causality analysis (GCA). Methods:Subjects aged 40 to 80 years, including 34 LA-MCI patients, 15 LA patients with normal cognition and 33 healthy controls, accepted resting-state functional magnetic resonance imaging. Independent component analysis was used to separate functional brain networks, and difference of activation was determined with two sample t-test. GCA was used to analyze effective connectivity of these functional networks. Results:Eight resting state networks were obtained, including default mode network, motor network, medial visual network, lateral visual network, right-memory network, left-memory network, auditory network and executive network. Activation was different among three groups. Effective connectivity of RSNs was also different among three groups. Conclusion:Components of the resting state networks keep changing as LA progressing. Activation decreases as patients' cognition impaired. The direction and strength of connections remodel.

Default memory network and working memory network in exercise addicts / 中华行为医学与脑科学杂志

Objective@#To explore the characteristics of the default memory network (DMN) and working memory network (WMN) at resting state brain functional network of exercise addiction people.@*Methods@#Twenty-nine sports addicts and 26 non-sports addicts matched by sex, age, average education level and sports dependence were screened by the exercise addiction index (EAI). Resting status brain scanning was performed with 3.0T magnetic resonance scanner.Sparse approximation coefficients independent component analysis (SACICA) model was used to analyze the independent components of brain networks.@*Results@#Compared with the DMN template, four features were extracted, including " basic conformity" , " less frontal lobe" , " more frontal lobe" and " less occipitoparietal lobe" . Compared with the parameters of " basic conformity" , the proportion of exercise addiction group (33.3%, 9/27) was higher than that of control group (18.2%, 4/22). In the other three parameters, the proportion of exercise addiction group (37.0%, 10/27; 3.7%, 1/27; 22.2%, 6/27) was lower than those of control group (45.5%, 10/22; 22.7%, 5/22; 27.3%, 6/22). But Chi-square test showed that there was no significant difference between the two groups(all P>0.05). Compared with the WMN template, six features were extracted, including " basic conformity" , " more frontal and parietal lobes" , " more parietal lobes" , " more frontal lobes" , " less frontal lobes" and " less parietal lobes" . The percentages of the first three features in exercise addiction group (22.2%, 6/27; 7.4%, 2/27; 7.4%, 2/27) were less than those in the control group (45.5%, 10/22; 22.7%, 5/22; 9.1%, 2/22), while the percentages of the last three features in the exercise addiction group (7.4%, 2/27; 37.0%, 10/27; 14.8%, 4/27) were higher than those in the control group (4.5%, 1/22; 13.6%, 3/22; 0, 0). Chi-square test showed that there was no significant difference in all features between the two groups was statistically(P>0.05).@*Conclusion@#No significant characteristic changes are found in DMN and WMN networks of exercise addiction population.

Investigation of default mode networks in rhesus macaque brain with 7.0T resting-state fMRI / 中国医学影像技术

Objective To investigate default mode networks (DMN) in healthy rhesus monkey brain.Methods Under anesthesia,two healthy rhesus macaques underwent resting state fMRI at a 7.0T MR scanner.The functional images were firstly normalized to the standard rhesus monkey template 112SM-RL-T1,and the GIFT software was aplied to conduct group-level independent component analysis on all preprocessed functional images.Results The functional connectivity maps of the resting-state networks of rhesus macaques were obtained using this method.DMN bilaterally included posterior cingulate cortex,anterior cingulate cortex,medial parietal cortex,retrosplenial cortex,arcuate sulcus,ventral intraparietal area,temporoparietal area,superior temporal sulcus dorsal bank and a portion of visual cortex.Conclusion With the help of cutting-edge 7.0T fMRI technology,It was demonstrated that DMN of rhesus macaque brain highly resembled the ones in human,which could support the notion that non-primates are useful models for neuropharmacological and neurocognitive studies.

Effects of the abnormalities in functional connectivity of the affective network on the relapse of major depressive disorder / 实用放射学杂志

Objective To explore abnormalities in functional connectivity of the affective network (AN) in relapse of major depressive disorder (MDD) after antidepressant treatment combined with resting state functional connectivity analysis.Methods Eleven recurrent MDD subjects after treatment,seventeen non recurrent MDD subjects after treatment and seventy-two healthy controls underwent fMRI scan.The amygdala,the pallidum,the insular cortex and the anterior cingulate cortex of the AN were selected as the template.Group independent component analysis (ICA) was performed to decompose the fMRI images into spatially independent components and the independent component which fit this template best was selected as AN.Two-sample t-tests were performed to investigate the changes in functional connectivity of the AN.Finally,the right amygdala and the medial prefrontal cortex were defined as seed regions.Results Compared with healthy control subjects and non-recurrent MDD group,recurrent MDD group showed significantly increased functional connectivity in the right amygdala in AN(P＜0.001).Meanwhile,the functional connectivity between the right amygdala and the medial prefrontal cortex was significantly decreased in recurrent MDD group(P ＜0.05).Conclusion Abnormal resting-state functional connectivity of the right amygdala after antidepressant treatment in MDD was found,suggesting that altered amygdala functional connectivity may serve as a predicator of relapse of the MDD.

Systematic Review of the Utility of Functional MRI to Investigate Internet Addiction Disorder: Recent Updates on Resting State and Task-Based fMRI

@#Introduction: Internet addiction disorder (IAD) particularly the internet gaming disorder (IGD) is recognized as a type of addiction similar to substance abuse. This addiction carries similar social impact as the latter, as it can cause serious impairment of interpersonal relationship, and even deterioration of academic or occupational performances. Functional magnetic resonance imaging (fMRI) is able to act as a non-invasive objective biomarker to detect functional neuronal connectivity in areas of the brain affected by IAD by utilizing blood oxygenation level dependent (BOLD) imaging. Methods: A systematic review was conducted from original articles published from January 2014 to January 2017 that had the keywords “internet addiction” and fMRI. Results: Initial data collection had 170 articles, however after applying the inclusion and exclusion criteria, there were 34 articles in the final analysis (17 resting-state fMRI studies and 18 task-based fMRI studies). The striatal nucleus and dopaminergic system demonstrated impaired functioning in subjects with IAD. Conclusion: Task-based and resting-state fMRI are able to detect areas of the brain that are activated in subjects with internet addiction, similar to those observed in subjects with substance abuse and other addictions. This review also introduces a newly arising subtype which is smartphone addiction disorder.

Efecto del pre-procesamiento del EEG en el Análisis por Componentes Independientes: reducción del artefacto del implante coclear en los Potenciales Evocados Auditivos / Effect of EEG pre-processing on Independent Component Analysis: reduction of cochlear implant artifact in Auditory Evoked Potentials

Resumen: El Análisis por Componentes Independientes (ICA, Independent Component Analysis) es una herramienta muy utilizada para eliminar los artefactos comunes del EEG, sin embargo existe poca bibliografía sobre el impacto que tienen etapas de pre-procesamiento de esta señal sobre el desempeño del ICA. En este trabajo se comparó el efecto de aplicar dos filtros digitales diferentes, pasabajas y pasabanda, en una etapa de procesamiento previa a ICA, para remover específicamente el artefacto de un implante coclear en registros de Potenciales Evocados Auditivos. Se analizaron señales de 10 sujetos usuarios de implante coclear y en 5 de estos registros con el pre-filtrado pasabajas se obtuvieron los mayores valores del índice de la Relación Señal Interferencia, utilizado para evaluar la calidad de la separación. El mayor efecto al remover el artefacto del implante coclear se nota en los electrodos T4 y T6, que corresponde a la zona donde los sujetos tienen colocado su implante (área temporal derecha).

Abstract: Independent Component Analysis (ICA) is an algorithm used to remove artifacts from the EEG. However, there is little current literature about the impact of preprocessing stages of this signal on the performance of ICA. In this paper the effect of applying two different digital filters - lowpass and bandpass -, in a pre-processing step to ICA, was compared. This to remove the cochlear implant artifact from the Auditory Evoked Potentials. Recordings from 10 cochlear implant users were analyzed. In 5 of these records using the pre-lowpass filtering, the highest Signal Interference Ratio (SIR) was obtained; this index was used to assess the quality of ICA separation. The greatest effect of removing the cochlear implant artifact is noted in both T4 and T6 electrodes, which correspond to the area where the subjects have placed their implants (right temporal area).

Dysconnectivity of multiple resting-state networks in patients with schizophrenia with auditory verbal hallucinations / 中国心理卫生杂志

Objective:To explore the traits of functional connectivity of resting-state networks in patients of schizophrenia with auditory verbal hallucinations (AVH) by using independent component analysis (ICA).Methods:All patients were met the Diagnostic and Statistical Mannal of Mental Disorders,Fourth Edition,Text Revision (DSM-IV-TR) diagnostic criteria for schizophrenia.Thirty schizophrenia patients with frequent AVH (AVH),24 schizophrenia patients without AVH (non-AVH) and 60 healthy controls (HC) matching with age and gender were analyzed by resting functional magnetic resonance imaging.The AVH were assessed by using Hoffman auditory hallucination scale.By using ICA,auditory network (AUN),language network (LAN),salience network (SAN),right front-parietal network (RFP) and default mode network (DMN) were selected as interesting networks for further analyses.Covariance analysis was used to compare the activity and functional connectivity of multiple resting state networks between three groups,and correlation analysis was used to evaluate the relations between abnormalities of brain and the severity of hallucination.Results:Compared to non-AVH,AVH group showed increased activity in left superior temporal gyrus (STG) and left postcentral gyrus (LPG) in the AUN,and showed decreased activity in right anterior cingulate cortex (ACC) in the SAN (P ＜ 0.05,FDR corrected).The severity of AVH group was associated with activity of the left STG (r =0.43,P ＜0.05) and the right ACC (r =-0.48,P ＜0.01).Furthermore,compared to non-AVH group,there were decreased connectivity between LAN and AUN [(-0.29 ±0.21) vs.(-0.16 ± 0.17),P ＜ 0.05,uncorrected],decreased connectivity between SAN and LAN [(-0.09 ± 0.22) vs.(-0.06 ± 0.18),P ＜ 0.05,uncorrected],increased connectivity between SAN and AUN [(0.30 ± 0.18) vs.(0.15 ± 0.24),P ＜ 0.05,uncorrected] in AVH group.Conclusion:There may be aberrant functional connectivities of AUN,LAN and SAN in schizophrenia patients with frequent AVH,and the occurrence of AVH seems to be related to brain areas involved in language production,speech perception and self-monitoring.

Diabetes classification using a redundancy reduction preprocessor

Introduction Diabetes patients can benefit significantly from early diagnosis. Thus, accurate automated screening is becoming increasingly important due to the wide spread of that disease. Previous studies in automated screening have found a maximum accuracy of 92.6%. Methods This work proposes a classification methodology based on efficient coding of the input data, which is carried out by decreasing input data redundancy using well-known ICA algorithms, such as FastICA, JADE and INFOMAX. The classifier used in the task to discriminate diabetics from non-diaibetics is the one class support vector machine. Classification tests were performed using noninvasive and invasive indicators. Results The results suggest that redundancy reduction increases one-class support vector machine performance when discriminating between diabetics and nondiabetics up to an accuracy of 98.47% while using all indicators. By using only noninvasive indicators, an accuracy of 98.28% was obtained. Conclusion The ICA feature extraction improves the performance of the classifier in the data set because it reduces the statistical dependence of the collected data, which increases the ability of the classifier to find accurate class boundaries.

On the Identification of an ICA Algorithm for Auditory Evoked Potentials Extraction: A Study on Synthetic Data / Identificación de un algoritmo ICA para la extracción de potenciales evocados auditivos: un estudio en señales sintéticas

Extracting characteristics and information from Auditory Evoked Potentials recordings (AEPs) involves difficulties due to their very low amplitude, which makes the AEPs easily hidden by artifacts from physiological or external sources like the EEG/EMG, blinking, and line-noise. To tackle this problem, some authors have used Independent Component Analysis (ICA) to successfully de-noise brain signals. However, since interest has been mainly focused on removing artifacts like blinking, not much attention has been paid to the quality of the recovered evoked potential. This is the AEP case, where literature reports interesting results on the de-noising matter, but without an objective evaluation of the AEP finally extracted (and the influence of different implementations or configurations of ICA). Here, to study the performance of three popular ICA algorithms (FastICA, Ext-Infomax, and SOBI) at separating AEPs from a mixture, a synthetic dataset composed of one Long Latency Auditory Evoked Potential (LLAEP) signal and the most frequent artifacts was generated. Next, the quality of the independent components (ICs) estimated by such algorithms was measured by using the AMARI performance index (Am), the signal interference ratio index (SIR), and the time required to achieve separation. Results indicated that the FastICA implementation, with the symmetric approach and the power cubic contrast function, is more likely to provide the best and faster separation of the LLAEP, which makes it suitable for this purpose.

La extracción de características e información de los registros de Potenciales Evocados Auditivos (AEPs) es complicada debido a su baja energía, la que lo hace fácilmente enmascarable por artefactos de origen fisiológico o externo, como el EEG/EMG, el parpadeo y el ruido de línea. Este problema ha sido abordado por algunos autores mediante el uso del Análisis por Componentes Independientes (ICA, por sus siglas en inglés), que se ha utilizado principalmente para reducir artefactos. Estos trabajos han enfocado su interés en la tarea de remover artefactos como el parpadeo, por lo que han descuidado el estudio de la calidad del potencial evocado recuperado. Este es el caso del AEP, donde aun cuando la literatura reporta resultados interesantes en la reducción de artefactos, no existe una evaluación objetiva del AEP finalmente extraído (y el efecto de usar diferentes implementaciones/configuraciones de ICA). En este trabajo, con el objetivo de cuantificar el desempeño de tres algoritmos de ICA (FastICA, Ext-Infomax, y SOBI) en la calidad de la separación de los AEPs, se generó una mezcla sintética de señales compuesta por un Potencial Evocado Auditivos de Latencia Larga (LLAEP) y artefactos frecuentemente presentes en estos registros. Después, se cuantificó la calidad de los componentes independientes (ICs, por sus siglas en inglés) estimados por estos algoritmos utilizando el índice de desempeño (AMARI, por sus siglas en inglés) el índice de la relación de interferencia entre señales (SIR, por sus siglas en inglés) y el tiempo requerido para realizar la separación. Los resultados indican que FastICA, con el enfoque simétrico y la función de contraste potencia cúbica, proporciona la mejor y más rápida separación del LLAEP, lo que lo vuelve idóneo para esta tarea.

Default Mode Network Altered in Chronic Pain Caused by Cervical Spondylosis / 中国康复理论与实践

Objective To explore the changes of resting state default mode network (DMN) in patients with chronic pain caused by cer-vical spondylosis (CPCS). Methods 8 healthy controls and 10 patients accepted functional MRI scanning. Surface based DMN was extract-ed with independent component analysis (ICA). The functional connectivity of the components of DMN were discriminated with support vector machine (SVM) algorithm from the patients to the controls. Results The DMN connectivity was different in the patients from the con-trols in some of the component areas. Conclusion DMN of CPCS patients is disorder in multiple brain areas, which may be involved with dysfunction of perception processing, emotion and memory.

A Network Analysis of 15O-H2O PET Reveals Deep Brain Stimulation Effects on Brain Network of Parkinson's Disease

PURPOSE: As Parkinson's disease (PD) can be considered a network abnormality, the effects of deep brain stimulation (DBS) need to be investigated in the aspect of networks. This study aimed to examine how DBS of the bilateral subthalamic nucleus (STN) affects the motor networks of patients with idiopathic PD during motor performance and to show the feasibility of the network analysis using cross-sectional positron emission tomography (PET) images in DBS studies. MATERIALS AND METHODS: We obtained [15O]H2O PET images from ten patients with PD during a sequential finger-to-thumb opposition task and during the resting state, with DBS-On and DBS-Off at STN. To identify the alteration of motor networks in PD and their changes due to STN-DBS, we applied independent component analysis (ICA) to all the cross-sectional PET images. We analysed the strength of each component according to DBS effects, task effects and interaction effects. RESULTS: ICA blindly decomposed components of functionally associated distributed clusters, which were comparable to the results of univariate statistical parametric mapping. ICA further revealed that STN-DBS modifies usage-strengths of components corresponding to the basal ganglia-thalamo-cortical circuits in PD patients by increasing the hypoactive basal ganglia and by suppressing the hyperactive cortical motor areas, ventrolateral thalamus and cerebellum. CONCLUSION: Our results suggest that STN-DBS may affect not only the abnormal local activity, but also alter brain networks in patients with PD. This study also demonstrated the usefulness of ICA for cross-sectional PET data to reveal network modifications due to DBS, which was not observable using the subtraction method.