Background
Immunohistochemistry is a routine method for evaluating the distribution of antigens in microscopic tissue samples. For example, visualization of smooth muscle actin in histology is commonly used to study vascular physiology, understand disease progression and/or visualize organ development. To achieve this, the distribution of smooth muscle actin in the tissue of interest is quantified to reveal the overall number of vessels present, and/or determine area, perimeter and thickness of the vasculature. Determining the distribution of proteins in tissue samples is performed using manual and semi-automatic methods. Manual methods involve sampling small areas and then extrapolating the results to the whole tissue — a method of estimation that is prone to error. Alternatively, several semi-automatic approaches exist that capitalize on a computer's ability to determine protein distributions by mapping based on colour thresholds (e.g., selecting everything that stains a certain brown colour). However, the effectiveness of this approach is undermined by the variability in staining (colour) intensity present between different sections of the tissue. Furthermore, these semi-automated techniques lack a method to combine adjacent tissue samples to provide a broader 3D context in larger tissue samples.
Technology Overview
The present invention allows for a highly accurate 3D reconstruction of tissue using a novel segmentation technique that utilizes an adaptive extraction method to localize vessel boundaries. In doing so, it provides a fully automated means to determine vessel area, perimeter and thickness. Most importantly, this invention provides a valuable tool for high-throughput vascular segmentations and morphological measures for analyzing and comparing disease-states. When compared against the manual segmentations, the automated segmentations demonstrated reliable results in significantly less time.
Benefits/Applications
• Provides the ability for high-throughput analysis
• Allows for automated segmentation and quantification.
• Measures vascular morphology and density, leading to complete accurate and repeatable description of smooth muscle in the microvasculature
• Excellent concordance with manual measurement
• Easy integration as a "plug-in" into an already existing image viewing workstation
Keywords
3D reconstruction, immunohistochemistry, automated segmentation, tissue samples, vascular morphology, vascular density, high-throughput, disease states
Patent Status: US 15/553,842 pending
