Analysis And Assessment Of Risk Factors For Mild Cognitive Impairment Based On The NIRS Brain Functional Network

With the ageing of the world population,the number of patients with mild cognitive impairment(MCI)is increasing year by year;this increase has seriously affected the quality of life of the patients and their families.At present,an accurate,scientific,and effective assessment of MCI remains a challenge.Thus far,several researchers have reported that hypertension,sleep,and fatigue are the risk factors of cognitive dysfunction,but the mechanism of these factors leading to cognitive dysfunction is not clear.Near-infrared spectroscopy(NIRS)can noninvasively and continuously measure the changes in the local oxyhemoglobin(?[HbO2])concentrations in the cerebral cortex.This technique has several potential advantages in terms of non-invasion,portability,low cost,and convenient operation.In this study,NIRS was used to build an evaluation modelbased on the brain functional connectivity(FC)and effective connectivity(EC)networks.We applied this model to study the characteristics of the brain functional network of the MCI population and to analyze the effects of the abovementioned risk factors(hypertension,sleep,and fatigue)on the brain connectivity network.On the basis of the research results,we conducted the preliminary development of a cognitive rehabilitation training system.In this study,first,the blood pressure index,sleep index,fatigue index,and cognitive level index of the MCI population were obtained by using the corresponding scale tests,and the relationship between these indices was determined using regression analysis.An evaluation model based on the brain FC and EC networks was established using the NIRS technique to assess the characteristics of the brain connectivity network of the MCI population.The phase synchronization between the brain regions calculated using wavelet phase coherence(WPCO)was applied to describe the brain FC.The coupling strength(CS)and coupling direction between channels calculated using Bayesian inference was applied to describe the EC.A correlation analysis between the characteristics of the brain connectivity network and the cognitive level was performed.The results showed that the WPCO and the CS values of the MCI group in intervals III and IV were significantly lower than those in the healthy controls.The results of the correlation analysis revealed that the characteristics of the brain connectivity network had a significantly positive correlation with the cognitive performance in the MCI group.These results demonstrated that the NIRS-based model of the brain connectivity network could be used to quantitatively assess the cognitive level of the subjects.Second,to analyze the relationship between the risk factors of MCI and the cognitive function,the NIRS measurement was used to analyze the characteristics of the brain connectivity network for hypertension,sleep disorders,and fatigue.(1)Effects of hypertension on brain FC and EC.In all,14 NIRS channels distributed in the bilateral prefrontal,sensorimotor,and occipital lobes were used to measure the cerebral ?[HbO2]concentrations of all of the brain regions of patients with hypertension and of the healthy controls.The FC and EC values were calculated after the data pre-processing.The results showed that both the WPCO and the CS values of the patients with hypertension significantly decreased in intervals III and IV,as compared to those of the healthy controls.Pearson's correlation analysis showed that both WPCO and CS had a significantly positive correlation with the cognitive performance in patients with hypertension.(2)Effects of sleep disorders on brain FC and EC.The cerebral ?[HbO2]signals of subjects with sleep deprivation and sleep disorders were recorded using the NIRS equipment.The FC and EC values between the brain regions were calculated after data pre-processing by using the moving average method and a Butterworth filter.The results showed that the WPCO and CS values after sleep deprivation exhibited a decreasing trend as compared to those in the control state.The ?[HbO2]signals were recorded for the sleep disorders group and the healthy control group by using the NIRS measurement.The FC and EC values were calculated in the frequency interval of 0.01-0.08 Hz,which represented the physiological meaning of spontaneous neural activity.The results showed that the sleep disorders group exhibited a decreasing trend in WPCO and CS as compared to the healthy controls.(3)Effects of fatigue on brain FC and EC.The results of the assessment of the characteristics of the brain connectivity network of subjects with chronic fatigue(seafarers)showed that the WPCO values of seafarers with long-term operation on the sea in intervals ?,?,and ? significantly decreased as compared to those in the healthy controls.The CS values of the seafarers group in intervals ? and ? were significantly lower than those of the healthy controls.Finally,on the basis of the characteristics of the brain connectivity network of the patients with MCI and the population with the risk factors for MCI,we combined the theoretical knowledge with practical applications,and designed and improved the following two cognitive tasks:a memory judgment task based on digital technology and a simulation driving task based on virtual reality.The ultimate goal was to maintain or improve the cognitive function of the trainers through the implementation of different cognitive tasks.In summary,an evaluation model based on the brain FC and EC networks was established to reflect the characteristics of the brain connectivity network.Furthermore,this evaluation model was used to assess the population with MCI or the risk factors for MCI.A scientific cognitive training task was set up,and the human-computer interaction for the cognitive training system interface was designed on the basis of the evaluation results.