Jun 28, 2023
Automatic labeling tissue and cell of human skin
- 1Johns Hopkins University
Protocol Citation: Kyu Sang Han, Pei-Hsun Wu 2023. Automatic labeling tissue and cell of human skin. protocols.io https://dx.doi.org/10.17504/protocols.io.j8nlko5z5v5r/v1
License: This is an open access protocol distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Protocol status: In development
We are still developing and optimizing this protocol
Created: June 26, 2023
Last Modified: June 28, 2023
Protocol Integer ID: 84049
Funders Acknowledgement:
test
Grant ID: 12345678
Disclaimer
This is first upload from TMC - Johns Hopkins. Not intended for actual usage.
Abstract
This is first upload from TMC - Johns Hopkins
Tissue biopsy collection
Tissue biopsy collection
After punch/excisional biopsy from a operating room (OR), place the skin tissue into a tissue container prefilled with buffered formalin for 12-24 hours at room temperature (RT)
Discard formalin, rinse with PBS, refill PBS, and leave the tissue in PBS for 1 minute
Take tissue out on a large Petri dish and measure the tissue size
Label a slotted cassette with a pencil with tissue ID and place the tissue in the cassette
Put cassette with the tissue back into the tissue container filled with PBS
Paraffin embedding
Paraffin embedding
Dehydrate tissue by submerging cassettes into ethanol at increasing concentrations from 70% to 99%.
Embed cassette paraffin at 60 deg C and store at room temperature
Tissue sectioning
Tissue sectioning
Section tissue and place on water bath to expand
Place expanded tissue section on a Superfrost Plus microscope slide
Tissue scanning
Tissue scanning
Scan unstained microscope slide at 40X magnification using Hamamatsu nanozoomer S210
Converting image format from ndpi to ome.tiff
Converting image format from ndpi to ome.tiff
Use openslide library in python along with pyvips library to load whole slide image file in proprietary format (ndpi) from Hamamatsu and save as ome.tiff along with metadata.
Applying semantic segmentation model in MATLAB
Applying semantic segmentation model in MATLAB
Load the ome.tiff image using bioformat library for MATLAB and generate tiles to input into a deep learning model to semantically segment 11 different skin tissue compartments.