The Key Technique Of FNIRS In The Identification Of Patients With Depressive Disorders And Their Emotions

According to " the National Depression Blue Book 2022 ",approximately 1 billion people worldwide are suffering from mental disorders,and one person loses their life to suicide every 40 seconds,with low-and middle-income countries accounting for77% of global suicides.Especially after the New Coronary Pneumonia epidemics,the global burden of mental disorders has become heavier,with cases of Major Depressive Disorder(MDD)and anxiety disorders increasing by 28% and 26%,respectively,and a surge of 53 million patients with depression,an increase of 27.6%,posing a greater challenge to the diagnosis and treatment of depression.The clinical diagnosis of depression mainly relies on scale screening and subjective judgment of physicians,and there is no objective biological diagnostic index yet,which has problems such as low diagnostic efficiency and lack of unified objective assessment criteria.Functional NearInfrared Spectroscopy(fNIRS)can detect the temporal and spatial characteristics of brain activity through the attenuation of near-red light in brain tissue,which has made it popular with recent years for the study of depressive disorders and emotion processing.This has led to its popularity in recent years in the study of depressive disorders and emotion processing.In order to evaluate the clinical value of fNIRS technology in aiding the diagnosis of depressive disorders,and to explore the common physiological indicators for depression identification and emotional recognition,we have addressed four key issues in fNIRS depressive disorders research: simple methodology,insufficient depth,lack of "gold standard" for functional connectivity,missing information flows and lack of consistent physiological indicators.In order to address four key issues in the study of fNIRS depressive disorders,namely,the simplicity of the method,the lack of depth,the lack of the "gold standard" of functional connectivity,the lack of information flow,and the lack of consistency of physiological indicators,this paper addresses several key issues in fNIRS depressive disorder identification and emotional recognition from four levels,layer by layer.In this study,using a 22-channel fNIRS device,30 patients with depressive disorders and 30 healthy controls were selected to examine the hemoglobin concentration in the prefrontal cortex(PFC)of MDD patients under an emotional audio stimulation task.The findings provide a reliable technical method as well as reproducible physiological indicators of fNIRS for accurate screening and aiding diagnosis of depressive disorders.The main work and innovations are as follows:(1)In response to the problems of simple research methods and insufficient research depth in fNIRS depressive disorder studies,this paper constructs a depressive disorder identification and emotional recognition method based on spatio-temporal scales.From the perspective of spatio-temporal characteristics,we systematically assessed and compared the differences in blood oxygen concentration and brain area activation between patients with depressive disorders and healthy people under different moods,as well as the differences between each of them in different moods.The results suggest that the combined spatio-temporal scaled depressive disorder identification and mood identification study is an effective identification model,which not only avoids the effects of individual differences,but also can accurately obtain a series of differential results at the brain area level and more easily capture the subtle differences caused by blood oxygen changes.Based on this method,it was found that significant differences in blood oxygen concentrations were observed between MDD patients and healthy people across emotions,and that blood oxygen concentrations were significantly lower in the MDD group than in healthy controls under negative stimuli.These results confirmed that the prefrontal lobes of MDD patients have abnormal blood oxygen metabolic changes and cognitive processing abilities,and finally achieved an identification accuracy of 90%levels in patients with depressive disorders and 85% at the emotional level based on activation features.(2)To address the problem of lacking the "golden rule" of functional connectivity in fNIRS depressive disorders,In this paper,a reliable and effective functional connectivity analysis method was constructed from three coupling indicators: Coherence(COH),Pearson’s Correlation Coefficient(COR)and Phase Locking Value(PLV),respectively.This study found that MDD patients showed more significant connectivity edges in response to negative stimuli and that MDD patients were more likely to show an increase in connectivity strength in response to negative stimuli.By main effects analysis of these significant connectivity edges,the differential connectivity between the two groups of positive stimulation were mainly manifested in the right middle frontal gyrus,while under negative stimulation it was mainly distributed among the right dorsolateral superior frontal gyrus and the left middle frontal gyrus.The results indicated that the two populations showed inconsistencies in brain region topology and connection strength.By fusing different connectivity indicators,better identification results were obtained in depressive disorder identification and emotion identification work.(3)To address the problem of missing information flows in the study of effective connectivity in fNIRS depressive disorders,this paper proposes a combination of granger causality analysis and dynamic causality analysis,which provides a method of the first time for the field of effective connectivity studies in fNIRS and establishes a model of neural interactions in different brain regions.This method not only assesses the equilibrium mechanism of brain regions own pre-determined mechanisms,but also investigates the effects of changes in experimental conditions on information pathways.The results of the study showed that there were significant differences in the number of main effect causality between MDD patients and healthy population and that some left frontal connections of MDD patients showed bidirectional flow of information under negative stimuli.The effects of experimental conditions on information transmission of the right dorsolateral superior frontal gyrus to the left middle frontal gyrus were examined by constructing a submodel and a full model,respectively,through the method of model building.The results showed that positive stimulation inhibited the selfbalancing of the right dorsolateral superior frontal gyrus to a greater extent,and negative stimulation inhibited the connection from the right dorsolateral superior frontal gyrus to the left middle frontal gyrus in MDD patients,however,negative stimulation facilitated the above connection to a greater extent in the Healthy Control(HC)group.Therefore,the effect of negative stimulation on the status of the right dorsolateral superior frontal to left middle frontal gyrus connections can also be an important neurological indicator of MDD patients and the healthy population.(4)To address the problems of uncertainty of effective indicators in fNIRS depressive disorder recognition and the lack of effective indicators in a emotional recognition.In this paper,we explore the feature extraction methods of depressive disorder recognition and emotion recognition from two perspectives: statistical features and fNIRS vector features.The results show that the traditional statistical features cannot obtain satisfactory recognition results in recognition of depressive disorders,and the features cannot achieve the work of recognition of emotions.However,the fNIRS vectors feature showed excellent recognition ability in the recognition work,and with only the feature of Cerebral Blood Volume( ΔCBV),We obtained the recognition rate of depressive disorders at the level of 90%,and even achieved the classification accuracy of 94.95%in recognition of emotion in MDD group,and The fusion results in different vector features also showed a high level of classification results.We found that ΔCBV and ΔCBV+ |L| could not only serve as excellent markers for depressive disorder recognition,but also still serve as effective biomarkers for emotional recognition.In conclusion,this study extracted heterogeneous fNIRS bio-markers,which are important to a comprehensive understanding of depressive disorders and effective identification of emotion.