TY - JOUR
T1 - QuPath algorithm accurately identifies MLH1 deficient inflammatory bowel disease-associated colorectal cancers in a tissue microarray
AU - Porter, Ross
AU - Din, Shahida
AU - Bankhead, Peter
AU - Oniscu, Anca
AU - Arends, Mark J
N1 - Funding Information:
S.D. acknowledges the financial support of NHS Research Scotland (NRS) through NHS Lothian. No other author has a conflict of interest to declare. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.
Funding Information:
This research was funded by Edinburgh and Lothians Health Foundation (Lothian Health Board Endowment Fund, Scottish Registered Charity Number SC007342), grant number S03021, to R.J.P., M.J.A. and S.D.
Publisher Copyright:
© 2023 by the authors.
PY - 2023/5/28
Y1 - 2023/5/28
N2 - Current methods for analysing immunohistochemistry are labour-intensive and often 13 confounded by inter-observer variability. Analysis is time consuming when identifying small 14 clinically-important cohorts within larger samples. This study trained QuPath, an open-source 15 image analysis program, to accurately identify MLH1-deficient inflammatory bowel disease-16 associated colorectal cancers (IBD-CRC) from a tissue microarray containing normal colon and IBD-17 CRC. The tissue microarray (n=162 cores) was imunostained for MLH1, digitalised, and imported 18 into QuPath. A small sample (n=14) was used to train QuPath to detect positive versus no MLH1 19 and tissue histology (normal epithelium, tumour, immune infiltrates, stroma). This algorithm was 20 applied to the tissue microarray and correctly identified tissue histology and MLH1 expression in 21 the majority of valid cases (73/99, 73.74%), incorrectly identified MLH1 status in one case (1.01%), 22 and flagged 25/99 (25.25%) cases for manual review. Qualitative review found five reasons for 23 flagged cores: small quantity of tissue, diverse/atypical morphology, excessive 24 inflammatory/immune infiltrations, normal mucosa, or weak/patchy immunostaining. Of classified 25 cores (n=74), QuPath was 100% (95% CI 80.49, 100) sensitive and 98.25% (95% CI 90.61, 99.96) specific 26 for identifying MLH1-deficient IBD-CRC, 휿=0.963 (95%CI 0.890, 1.036) (p
AB - Current methods for analysing immunohistochemistry are labour-intensive and often 13 confounded by inter-observer variability. Analysis is time consuming when identifying small 14 clinically-important cohorts within larger samples. This study trained QuPath, an open-source 15 image analysis program, to accurately identify MLH1-deficient inflammatory bowel disease-16 associated colorectal cancers (IBD-CRC) from a tissue microarray containing normal colon and IBD-17 CRC. The tissue microarray (n=162 cores) was imunostained for MLH1, digitalised, and imported 18 into QuPath. A small sample (n=14) was used to train QuPath to detect positive versus no MLH1 19 and tissue histology (normal epithelium, tumour, immune infiltrates, stroma). This algorithm was 20 applied to the tissue microarray and correctly identified tissue histology and MLH1 expression in 21 the majority of valid cases (73/99, 73.74%), incorrectly identified MLH1 status in one case (1.01%), 22 and flagged 25/99 (25.25%) cases for manual review. Qualitative review found five reasons for 23 flagged cores: small quantity of tissue, diverse/atypical morphology, excessive 24 inflammatory/immune infiltrations, normal mucosa, or weak/patchy immunostaining. Of classified 25 cores (n=74), QuPath was 100% (95% CI 80.49, 100) sensitive and 98.25% (95% CI 90.61, 99.96) specific 26 for identifying MLH1-deficient IBD-CRC, 휿=0.963 (95%CI 0.890, 1.036) (p
KW - QUPath
KW - machine learning
KW - biomarket
KW - MLH1
KW - Colorectal cancer
KW - inflammatory bowel disease
KW - mismatch repair
KW - immunohistochemistry
KW - histology
U2 - 10.3390/diagnostics13111890
DO - 10.3390/diagnostics13111890
M3 - Article
SN - 2075-4418
VL - 13
JO - Diagnostics
JF - Diagnostics
IS - 11
ER -