Histological imaging and region of interest (ROI) annotation

SB Sara L. Best
YL Yuming Liu
AK Adib Keikhosravi
CD Cole R. Drifka
KW Kaitlin M. Woo
GM Guneet S. Mehta
MA Marie Altwegg
TT Terra N. Thimm
MH Matthew Houlihan
JB Jeremy S. Bredfeldt
EA E. Jason Abel
WH Wei Huang
KE Kevin W. Eliceiri
request Request a Protocol
ask Ask a question
Favorite

A bright field image of the entire H&E slide was collected with an Aperio CS2 Digital Pathology Scanner (Leica Biosystems) at 20× magnification. Each core of grade 1 and grade 4 was cropped to the size of 1520 pixels by 1520 pixels using Aperio ImageScope viewing software (Leica Biosystems). An expert in genitourinary pathology reviewed the cropped cores to re-confirm the grade information and selected the representative regions containing cancer cells and an adjacent stroma region with patterns consistent with Fuhrman grade the whole tumor had been categorized as. For each core, 2–3 ROIs with the size of 400 pixels by 400 pixels (202um by 202um) were annotated containing typical cancer cells of grade 1 or grade 4 while excluding confounding tissue features (i.e. adipose tissue, benign tissue). In total, 75 TMA cores were verified and annotated as grade 1 clear cell RCC and 55 TMA cores as grade 4 clear cell RCC. Each core represents an individual patient. As described below in the SHG imaging methods section, nine total cores were excluded for lack of signal.

Do you have any questions about this protocol?

Post your question to gather feedback from the community. We will also invite the authors of this article to respond.

post Post a Question
0 Q&A