Assessment Of Brain Networks And Prognostic Factors Analysis In Disorders Of Consciousness

【Background】Patients with disorders of consciousness（DOC）are characterized by alterations in arousal and/or awareness,and common causes of DOC include cardiac arrest,traumatic brain injury,cerebral hemorrhage,and ischemic stroke.Clinical treatments for DOC remain a challenge,and there are tens of thousands of patients who are facing severe treatment dilemmas in China.The management of DOC inevitably raises ethical concerns about the appropriate degree of life-sustaining treatment.Families suffer a significant financial and psychological burden because DOC cannot communicate and are dependent on others for care.Therefore,accurate outcome prediction is significant for both patients and their families.One of the main targets for the clinical management of DOC is to identify reliable prognostic markers that might effectively predict prognosis.Additionally,advances in neuroscience have built the evidence base for assessing consciousness following severe brain injuries.It has been repeatedly confirmed that the exclusive use of clinical consensus of behaviors resulted in high rates of misdiagnosis for unable to reflect the real consciousness level of DOC.To help solve the diagnostic and prognostic problems,a range of neuroimaging technologies have been developed.These methods are now being used to map patterns of residual brain function and structure,helping reduce diagnostic errors between the Vegetative state（VS）/Unresponsive wakefulness syndrome（UWS）and Minimally conscious state（MCS）,and improving outcome prediction.Functional near-infrared spectroscopy（f NIRS）is a burgeoning optical brain monitoring technique that can measure the blood oxygenation level-dependent hemodynamic responses to neural activations.It has emerged as a promising technique in the field of DOC by virtue of its being non-invasive,portable,and low-cost.Several studies with small sample sizes have used task-based f NIRS to detect residual brain function in DOC.Compared with detecting task-oriented brain activities,resting-state f NIRS reflects spontaneous brain network properties and has the advantages of simplicity,convenience,and time-saving.However,there are no studies using resting-state f NIRS to explore the task-negative brain network.Thus,how hemodynamic changes occur in the cerebral cortex in the resting state in DOC is yet unknown.【Objectives】1.To analyze the factors related to long-term clinical outcomes in patients with prolonged disorders of consciousness（p DOC）and construct a simple and convenient nomogram model2.To investigate the topological structure and strength of spontaneous network interactions in the prefrontal cortex of individuals with different levels of consciousness,and to assess the discriminative power of f NIRS for VS/UWS and MCS.To identify the cerebral regions and connections critical for consciousness preservation and eligible for differentiating conscious states.【Methods】1.From January 2014 to January 2021,patients with VS/UWS or MCS were enrolled at the neuro-intensive care unit of Xijing Hospital.We examined the differences between the patients with a favorable prognosis（GOSE>3）and those with an unfavorable prognosis（GOSE≤3）.The clinical characteristics were collected retrospectively.Variables with P<0.05 in univariate analysis were included in a multivariate Logistic model.An outcome prediction nomogram was constructed based on the variables with P<0.05 in the multivariate analysis.2.In this study,all consecutive DOC patients admitted to the neuro-intensive care unit of Xijing Hospital from March 2021 to November 2021 were prospectively included.We collected clinical data from DOC patients and performed a resting-state f NIRS examination.Ten regions of interest（ROIs）in the prefrontal cortex（PFC）were selected:both sides of Brodmann area（BA）9,BA10,BA44,BA45,and BA46.Graph-theoretical analysis and seed-based correlation analyses were used to investigate the network topology and the strength of pairwise connections between ROIs and channels.The areas under the receiver operating characteristic curves of f NIRS features were examined to determine their discriminative power for UWS and MCS.【Results】1.From January 2014 through January 2021,a total of 170 patients with p DOC were admitted to our hospital.After screening via inclusion and exclusion criteria,a total of 55patients（41 men,14 women）were included in the final analysis.Univariate analysis showed there were significant differences in EEG background（P=0.004）and cortex metabolic status（P<0.001）among VS/UWS,MCS-and MCS+.Consciousness state,etiology,Coma recovery scale-revised score,EEG background were shown to be independent predictors for awareness recovery.The model demonstrated good discriminative power with an area under the receiver operating characteristic curve of 0.87.2.We performed resting-state f NIRS in 24 healthy controls and 23 DOC patients of whom12 were in MCS and 11 were in VS/UWS.We found that MCS and VS/UWS exhibited varying degrees of destruction of topological architecture,and the regional nodal properties of BA10 were significantly different between them(Nodal degree,P_Left BA10=0.01,P_Right BA10<0.01;nodal efficiency,P_Left BA10=0.03,P_Right BA10<0.01).Compared to healthy controls,VS/UWS had impaired functions in both short-and long-distance connectivity,however,MCS had significantly impaired functions only in long-distance connectivity.The functional connectivity of right BA10（AUC=0.88）and the connections between left BA46 and right BA10（AUC=0.86）had excellent performance in differentiating MCS and VS/UWS.【Conclusions】1.A nomogram risk prediction model for p DOC patients was established with good prognostic ability and compliance.This multimodal prognostic model can be used to identify patients with a higher likelihood of clinical improvement and could provide a reference for the prognosis of clinical outcome.2.MCS and VS/UWS have different patterns of topological architecture and short-and long-distance connectivity in PFC.Intra-connections within right BA10 and inter-hemispheric connections between right BA10 and left BA 46 are excellent resting-state f NIRS classifiers for distinguishing between MCS and VS/UWS.