Node degree distribution in complex microvascular networks: a potential new diagnostic tool for extracellular matrix-related diseases.

Recent research indicates the presence of increased vascular density and irregularity on oral mucosal vascular networks in extracellular matrix (ECM)-related illness or conditions. Here, we estimated the frequency of occurrence of nodes of various degrees (K3, K4 and K5, where Kn designates a node with n connections) in patients with proven or suspected ECM-related conditions and in controls. Subjects with ECM-related conditions exhibited lower K3 and higher K4 frequency than controls (p < 0.0001) in their vascular networks. Inverse statistical correlations between the local fractal dimension and L-Z values and percentage of K3 (Pearson's r values range: -0.91 to -0.81; p values range: 0.0013 to < 0.0001), together with a positive relationship with K4 were observed (r values range: 0.81 to 0.86; p values range: 0.0015 to 0.0003). A positive correlation coefficient between D(1–46) and K5 frequency was also found (r = 0.6334, p = 0.027). K3 ≥ 52% or K4 <28% discriminated ECM patients from controls with 100% sensitivity (true positive cases to true positive + false negative ratio) and specificity (true negative cases to true negative + false positive ratio). These findings suggest that node degree distribution in oral vascular networks could be a helpful new marker of pathological conditions associated with proven or suspected ECM abnormalities.