Multimodal Phasor Analysis for Digital Pathology: Quantitative Characterization of Liver Iron Overload

Area of Science:

• Digital Pathology
• Computational Pathology
• Medical Image Analysis

Background:

• Hemochromatosis involves excessive iron storage, primarily in the liver, potentially causing severe health issues.
• Current diagnosis relies on visual assessment of Perls' Prussian Blue (PB) stained tissue, a semiquantitative method prone to subjectivity.
• Whole-slide imaging (WSI) enables digital analysis of tissue sections, offering opportunities for quantitative diagnostics.

Purpose of the Study:

• To develop and validate a WSI processing method using a multimodal phasor approach for quantifying iron deposits in hemochromatosis.
• To provide objective, quantitative data on iron extent, size, and distribution, correlating with established diagnostic scoring systems.

Main Methods:

• A multimodal phasor approach was employed to process RGB whole-slide images (WSIs) of PB-stained tissue.
• Discrete Fourier Transform mapped spectral data to the phasor plane, enabling isolation of PB stain via unsupervised clustering.
• Phasor analysis of spatial autocorrelation determined the size of Prussian Blue aggregates.

Main Results:

• The developed method quantitatively assessed iron deposit extent, size, and distribution.
• Quantitative data generated by the phasor approach showed strong correlation with pathologists' diagnoses using the Scheuer scoring system.
• Phasor-based analysis demonstrated efficiency in segmentation and management of large WSI datasets.

Conclusions:

• Phasor-based image analysis offers a robust, quantitative method for assessing iron accumulation in hemochromatosis from WSIs.
• This approach enhances diagnostic accuracy and efficiency, supporting pathologists in clinical decision-making.
• The method holds potential for analyzing other spectral imaging applications, including immunohistochemistry and nanoparticle studies.