https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11109499/
Skin Res Technol. 2024 May; 30(5): e13763. Published online 2024 May 21. doi: 10.1111/srt.13763
Sara Bohjanen, Brian McAdams, Connor Goldstick, and Maria Hordinsky
1. RESEARCH LETTER
The role of mast cells in allergic inflammation has been long‐established, but the involvement of mast cells in other inflammatory and autoimmune diseases is a newer and expanding area of research. For example, mast cells have been implicated in the disease processes of atopic dermatitis, 1 psoriasis, 1 , 2 and a variety of other conditions. 3 , 4 Studies examining the contribution of mast cells to disease pathogenesis and analyzing the efficacy of mast cell‐targeted therapies will require extensive quantification of mast cells and their granules. We propose a method to facilitate this type of analysis by utilizing machine learning (ML) to automatically detect mast cells and granules in skin biopsy specimens imaged with confocal microscopy.
Twenty‐two punch biopsy specimens originating from a pilot study 5 were used in this study. This included lesional and perilesional skin from six subjects with plaque psoriasis (four males, two females, age range of 37–53 years old). The specimens were cryosectioned into 60‐µm‐thick vertical sections and immunostained with mouse anti‐tryptase antibody and fluorophore‐conjugated secondary antibodies for mast cell visualization. Images of the sections were acquired with confocal microscopy, and then they were processed, labeled, and used to train and validate the ML model 6 (see Supplementary Material). The ML model was subsequently used to calculate cell and granule densities (per mm2) of the biopsy specimens. Linear regressions were performed to examine the relationship between mast cell (or granule) density and the presence of plaque psoriasis (e.g., lesional vs. perilesional specimen), as well as depth of subsection. All statistical analyses were performed in R (R version 4.2.2). The study was reviewed and approved by the Institutional Review Board of the University of Minnesota (STUDY00015853).
[More]
 |
Virus-free.www.avast.com |
Mastopedia Research Feed 