An Integrated Multiphoton Imaging Workflow for Quantitative Analysis of Aortic Tissue Microstructure

Quantitative characterization of aortic microstructure is essential for advancing vascular biomechanics and mechanobiology. To address this need, we present an image-analysis workflow that extracts microstructural descriptors from multiphoton microscopy of the murine descending thoracic aorta. The workflow is demonstrated in a representative vessel imaged under physiological loading. Channel-specific signals are acquired for collagen (second harmonic generation), elastin (two-photon autofluorescence), and cell nuclei (two-photon excited fluorescence). Following reorientation into the through-thickness plane, elastic lamellae are traced to quantify thickness and interlamellar spacing using circle-based geometry. After correcting for vessel wall curvature using a cylindrical transformation, segmented nuclei are assigned to media or adventitia based on the position of their centroid relative to the boundary between the two layers, identified in the flattened vessel representation; nuclear morphology is then characterized using an inertia-tensor-based equivalent ellipsoid to quantify nuclear aspect ratio and major-axis orientation. Collagen organization is characterized from optical sections by extracting fiber centerlines to quantify straightness and amplitude; fiber traces from serial sections are then stacked to generate an approximate three-dimensional representation, from which apparent porosity and linear fiber density are estimated, while fiber-orientation distributions derived from principal component analysis are characterized using a von Mises mixture. Finally, collagen and elastin volume fractions are computed using a two-stage fixed-threshold approach, with thresholds established from a calibration subset of four additional vessels. Overall, this robust workflow provides a framework for studying aortic wall remodeling across physiological and pathological processes.