Fig 1.
Overview of image acquisition set-up and image processing.
(A) Schematic of microscope setup for imaging with linear polarization. (B) Demonstration of image acquisition scheme, where the same microscope slide was imaged at 6 angular rotations (0, 60, 120, 180, 240, and 300 degrees, numbered 1 through 6 above). After acquisition, the 6 images were co-registered and combined into a composite image using the maximum intensity at each pixel location. (C) Higher magnification image showing necrotic core (blue boundary) and the lumen (teal line) bounding the cap region (analysis is restricted to the cap region by this method). Polygons emanating from the lumen center point (yellow/red circle) at one degree intervals (dashed lines) are used for quantification of images for downstream analysis.
Table 1.
Summary of composite images and associated Δϴ and range of rotation angle (ϴR,full = 0° to 358°, ϴR,low = 0°to 178° and ϴR,high = 180°to 358°) for a single histology slide.
Fig 2.
(A) Four example serial sections (each cut at 7 μm thick) from the same 2 mm tissue block, and changes to both bright light (top) and polarized light (bottom) images are observed in the cap region. (B) Four different sections (numbered from left to right) from 3 different hearts showed increased repeatability in (C) with pixel counting method compared to use of grayscale intensity. Scale bars = 1 mm.
Fig 3.
Responsivity of pixel counting to changes in cap thickness.
(A) First and last of thirty serial sections (each cut at 7 μm thick), illustrating changing thickness of necrotic core and cap thickness at mid region of necrotic core arc (dotted white box), bright light images (top) and polarized images (bottom). Scale bars = 1 mm. (B) Plot of pixel counting method illustrating decreasing magnitude of positive cap pixels throughout the sections as the cap thickness decreases.
Fig 4.
Unpolarized image and corresponding polarized co-registered composite images from different Δϴ.
(A) First row and second rows correspond to an example of a section with a thick collagen cap (124μm) and thin cap (43μm), respectively. (B) Unpolarized and corresponding polarized composite images (Δϴ = 60°) of other samples for angle of rotation analysis. Scale bars = 1 mm.
Table 2.
Comparative results of total pixel intensity in the cap region for different Δϴ.
Table 3.
Comparative results of total pixel counts in the cap region for different Δϴ.
Table 4.
Results of total intensity in the cap region from composite images from half of the range of rotation angles compared to the full range.
Table 5.
Results of total pixel counts in the cap region from composite images from half of the range of rotation angles compared to the full range.
Fig 5.
Total number of collagen positive pixels in cap region for different threshold for pixel counting.
Values shown are relative to no threshold (intensity value = 0). (A) Composite image from Δϴ = 2° and ϴR,full and (B) Composite image from Δϴ = 60° and ϴR,full.