Hemoglobin Signal Network Mapping Reveals Novel Indicators for Precision Medicine

We introduce methodology, applied to a representative physiological time series measure (fNIRS) and disease type (breast cancer), that provides access to features more typical of molecular-cellular processes but here are observed from intact tissue. The methodology adopts a novel approach to generating a multivariate ordinal partition transition network to enable quantitative descriptions of short-term dynamics (<1sec). Accessed are several classes of adjacency matrices whose exploration and associated co-dependent behaviors unexpectedly reveals features of structured dynamics, some of which are shown to exhibit enzyme-like behaviors and sensitivity to recognized molecular markers of disease. Constituting class measures that are often easily applied and widely accessible, such sensitivities may afford novel extensions to on-demand phenotypic measures that have thus far had limited utility for complementing precision medicine strategies. Discussed are exemplary uses in support of this aim, and extension to other forms of physiological time series measures.