We previously proposed a novel taxonomic framework to describe the diffusion tensor imaging (DTI) profiles of white matter tracts by their diffusivity and neural properties. We have shown the relevance of this strategy toward interpreting brain tissue signatures in Classic Normal Pressure Hydrocephalus vs. comparator cohorts of mild traumatic brain injury and Alzheimer's disease. In this iteration of the Periodic Table of DTI Elements, we examined patterns of tissue distortion in Complex NPH (CoNPH) and validated the methodology against an open-access dataset of healthy subjects, to expand its accessibility to a larger community. DTI measures for 12 patients with CoNPH with multiple comorbidities and 45 cognitively normal controls from the ADNI database were derived using the image processing pipeline on the brainlife.io open cloud computing platform. Using the Periodic Table algorithm, DTI profiles for CoNPH vs. controls were mapped according to injury patterns. Structural volumes in most structures tested were significantly lower and the lateral ventricles higher in CoNPH vs. controls. In CoNPH, significantly lower fractional anisotropy (FA) and higher mean, axial, and radial diffusivities (MD, L1, and L2 and 3, respectively) were observed in white matter related to the lateral ventricles. Most diffusivity measures across supratentorial and infratentorial structures were significantly higher in CoNPH, with the largest differences in the cerebellum cortex. In subcortical deep gray matter structures, CoNPH and controls differed most significantly in the hippocampus, with the CoNPH group having a significantly lower FA and higher MD, L1, and L2 and 3. Cerebral and cerebellar white matter demonstrated more potential reversibility of injury compared to cerebral and cerebellar cortices. The findings of widespread and significant reductions in subcortical deep gray matter structures, in comparison to healthy controls, support the hypothesis that Complex NPH cohorts retain imaging features associated with Classic NPH. The use of the algorithm of the Periodic Table allowed for greater consistency in the interpretation of DTI results by focusing on patterns of injury rather than an over-reliance on the interrogation of individual measures by statistical significance alone. Our aim is to provide a prototype that could be refined for an approach toward the concept of a "translational taxonomy."
Read full abstract