The microstructure of hardened cement pastes comprises of a heterogeneous agglomeration of distinct quasihomogeneous domains with variable physical, chemical, and morphological features, denoted here as material phases. Accurate material characterization rests on a precise description and quantification of underlying principal phases, focusing, in particular, on their volumetric proportions and spatial configuration within the microstructure, as these affect, to a large extent, the macroscopic properties of the composite material. A realistic cement paste microstructure used in this study was obtained from micro-computed X-ray tomography, following the application of suitable segmentation filters, highlighting and isolating the sought phase – anhydrous cement grains – for statistical analysis. The present paper then compares and assesses several implementations of a lineal path function, all applied to quantifying the phase connectedness and short-range order characteristics of the grains. The main emphasis rests on assessing the accuracy and evaluation speed of the implemented algorithms.