Mild traumatic brain injury (mTBI) accounts for over one million emergency visits in the United States each year, leading to emotional, physical, and cognitive symptoms that significantly impact patients’ quality of life. While rsfMRI studies such as investigation of the default mode network (DMN) reported alterations in brain function (1-3), the heterogeneous nature of mTBI is a current major obstacle to characterizing these. Each patient is unique in terms injury location and biomechanical profile, making generalization of findings from group-level comparisons to an individual-level difficult. However, we hypothesize that mTBI patients may follow a similar pattern of alterations in higher order brain functions despite the heterogeneity of mTBI. To evaluate our hypothesis, we investigated the relationship between group-level and individual-level alterations in the DMN after mTBI using a novel approach, connectivity domain analysis.Seventeen (17) mTBI patients (age: 38.23±15.44 years) and 17 demographically matched healthy controls (age: 26.05±5.22 years) were recruited at the acute stage at Detroit Receiving Hospital. Figure 1 describes the analysis pipeline. For this study, anatomical information was used to identify the “seed networks,” and 145 regions of interest from Harvard-Oxford cortical and subcortical structural atlases were chosen. Pearson correlation was used as a connectivity index to calculate connectivity weights, the input data for the connectivity domain. Using the connectivity domain, the connectivity of the brain can be modeled among subjects and tested for differences across different studies and populations, which is one bottleneck in brain connectivity research. Next, a first-level feature-based ICA was applied to obtain resting state networks (RSNs) in the connectivity domain. The design matrix, or connectivity matrix (CM), represents the interaction between seed networks and RSNs. A one-sample nonparametric statistical test was performed on elements of the CM to identify the statistically significant elements, revealing the seed networks which are functionally associated with RSNs. Statistical nonparametric comparison was only performed on the seed networks associated with the DMN to reduce the multiple comparisons effect.The first-level feature-based ICA in the connectivity domain extracts similar RSNs as ICA analysis in the time domain, as demonstrated in Figure 2, comparing the DMN obtained from a common ICA analysis in the time domain with the DMN obtained by applying the feature-based ICA on the data in the connectivity domain. Applying a one-sample nonparametric statistical test, we identified the seed networks associated with RSNs. Figure 3 represents the seed networks associated with the DMN demonstrated in Figure 2. Seed networks positively interacting with the DMN include the angular gyrus (AG), lateral occipital cortex-superior division (OLs), cingulate gyrus-posterior division (CGp), precuneous cortex (PCN), supracalcarine cortex (SCLC), middle temporal gyrus-anterior division (T2a), and frontal medial cortex (FMC). Seed networks negatively interacting with DMN include precentral gyrus (PRG), postcentral gyrus (POG), supramarginal gyrus-anterior division (SGa), supplementary motor cortex (SMC), frontal orbital cortex (FO), insular cortex (INS), and inferior frontal gyrus-pars opercularis (F3o). A nonparametric statistical comparison performed at the group level revealed decreases of connectivity strength between the DMN and seed networks from both the areas directly associated with DMN, including CGp.L, PCN.L, T2a.R, FMC.R, and CGp.R, and the areas inversely associated with the DMN, including POG.L, PRG.R, and SGa.R (Figure 4). A subject-level analysis shows that although different seed networks show statistically significant differences across patients, in general, the same pattern was observed at the subject level as at the group level. Among seed networks which are positively associated with the DMN, 32 out of 34 statistically significant differences reveal a decrease in their strength. Similarly, 24 out of 26 statistically significant differences for seed networks that are inversely associated with the DMN reveal a decrease in their strength (Figure 5).The connectivity domain is a new framework for investigating brain functional activity and the interactions between brain networks at different levels of functional hierarchy. Similar to previous findings, we observed decreased activity within the DMN after mTBI. We also observed decreases in strength of connectivity of seed networks which are inversely linked to the DMN. Due to the heterogeneity of mTBI, a group view may distort the pathophysiological scenario of individual patient. In this study, we demonstrated that despite inter-individual variability, mTBI patients do share commons patterns of alterations in the DMN. Specifically, we demonstrated similar patterns of alteration in DMN activities among subjects despite different mechanisms and locations of the injury. This finding serves to validate previous group-level findings despite current disagreement toward them.This work suggests that in spite of heterogeneity of mTBI, similar patterns of DMN alterations can be observed across individuals.