Recently, DMN derived from coherent spontaneous blood oxygen level-dependent fluctuations in resting state functional MRI has been extensively evaluated and thought to represent the neural consciousness stream. Decreased DMN, especially of the posterior cingulate cortex, is associated with different degrees of impaired cognition and consciousness. Without requiring demanding tasks, rs-fMRI is suited for investigating cognitive and consciousness disorders in liver cirrhosis. The DMN is different between normal controls and HE patients. However, network alteration in patients without HE and those with minimal HE is unknown. The gap between cerebral edema, ammonia concentration, and functional network is still unsolved. The aim of the present study was to investigate connectivity in resting-state networks correlating with consciousness across several HE levels under the hypothesis that increased cerebral edema evaluated by DTI is associated with a loss of connectivity in DMN, as assessed by rs-fMRI functional connectivity. The BOLD signal from rs-fMRI is not a direct measure of neuronal activity but reflects local variations in de-oxyhemoglobin concentration from a combination of blood flow, blood volume, and oxygen metabolism. The results are consistent with a previous PET study with impaired blood flow and oxygen metabolism in the frontal cortices and anterior cingulate gyrus in liver cirrhosis. Changes in BOLD signal have been linked to synaptic activity of glutamate and its recycling through astrocytes from evidence on cell biology. By using rs-fMRI, disconnection of BOLD signal coherence was proved to correlate with the disease progression. Decreased fc-CC derived from the BOLD signal between intercortical regions may cause the underlying desynchronized neural activity of glutamate. Glutamate is the major metabolism of ammonia detoxification. However, we did not find significant association between venous ammonia level and functional connectivity in this study. Intervention for HE by lessening the ammonia concentration might alter results interpretation since ammonia level decrease can precede consciousness recovery. The altered glutamine-glutamate cycle can mediate many vital processes, including inNVP-BEZ235 PI3K inhibitor formation encoding, memory formation and retrieval, spatial recognition, and consciousness maintenance. Further animal studies should be conducted to validate these findings. The main neuro-pathologic findings in HE is astrocyte swelling and intra-cranial hypertension. Using DTI, interstitial edema is shown in HE in vivo. Altering glutamate re-uptake during ammonia metabolism may result in the intracellular depletion of myo-inositol, an organic osmolyte that can trigger macromolecule migration to the extracellular space, resulting in increased extracellular fluid accumulation. This accounts for the increased MD value. Another explanation for the elevated MD value may be the increased interstitial space from cell loss. The persistence of modified extra- and intra-cellular glutamate concentrations also alters the surrounding glial processes, while the associated mitochondrial dysfunction with oxidative stress may represent possible apoptosis with subsequent change in the brain network. However, the subtle FA decrease in the present study may only partially explain the hypothesis of cell loss.