Deep learning identifies inflamed fat as a risk factor for lymph node metastasis in early colorectal cancer
Publikation: Bidrag til tidsskrift › Tidsskriftartikel › Forskning › fagfællebedømt
Standard
Deep learning identifies inflamed fat as a risk factor for lymph node metastasis in early colorectal cancer. / Brockmoeller, Scarlet; Echle, Amelie; Ghaffari Laleh, Narmin; Eiholm, Susanne; Malmstrøm, Marie Louise; Plato Kuhlmann, Tine; Levic, Katarina; Grabsch, Heike Irmgard; West, Nicholas P.; Saldanha, Oliver Lester; Kouvidi, Katerina; Bono, Aurora; Heij, Lara R.; Brinker, Titus J.; Gögenür, Ismayil; Quirke, Philip; Kather, Jakob Nikolas.
I: Journal of Pathology, Bind 256, Nr. 3, 2022, s. 269-281.Publikation: Bidrag til tidsskrift › Tidsskriftartikel › Forskning › fagfællebedømt
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - JOUR
T1 - Deep learning identifies inflamed fat as a risk factor for lymph node metastasis in early colorectal cancer
AU - Brockmoeller, Scarlet
AU - Echle, Amelie
AU - Ghaffari Laleh, Narmin
AU - Eiholm, Susanne
AU - Malmstrøm, Marie Louise
AU - Plato Kuhlmann, Tine
AU - Levic, Katarina
AU - Grabsch, Heike Irmgard
AU - West, Nicholas P.
AU - Saldanha, Oliver Lester
AU - Kouvidi, Katerina
AU - Bono, Aurora
AU - Heij, Lara R.
AU - Brinker, Titus J.
AU - Gögenür, Ismayil
AU - Quirke, Philip
AU - Kather, Jakob Nikolas
N1 - Publisher Copyright: © 2021 The Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.
PY - 2022
Y1 - 2022
N2 - The spread of early-stage (T1 and T2) adenocarcinomas to locoregional lymph nodes is a key event in disease progression of colorectal cancer (CRC). The cellular mechanisms behind this event are not completely understood and existing predictive biomarkers are imperfect. Here, we used an end-to-end deep learning algorithm to identify risk factors for lymph node metastasis (LNM) status in digitized histopathology slides of the primary CRC and its surrounding tissue. In two large population-based cohorts, we show that this system can predict the presence of more than one LNM in pT2 CRC patients with an area under the receiver operating curve (AUROC) of 0.733 (0.67–0.758) and patients with any LNM with an AUROC of 0.711 (0.597–0.797). Similarly, in pT1 CRC patients, the presence of more than one LNM or any LNM was predictable with an AUROC of 0.733 (0.644–0.778) and 0.567 (0.542–0.597), respectively. Based on these findings, we used the deep learning system to guide human pathology experts towards highly predictive regions for LNM in the whole slide images. This hybrid human observer and deep learning approach identified inflamed adipose tissue as the highest predictive feature for LNM presence. Our study is a first proof of concept that artificial intelligence (AI) systems may be able to discover potentially new biological mechanisms in cancer progression. Our deep learning algorithm is publicly available and can be used for biomarker discovery in any disease setting.
AB - The spread of early-stage (T1 and T2) adenocarcinomas to locoregional lymph nodes is a key event in disease progression of colorectal cancer (CRC). The cellular mechanisms behind this event are not completely understood and existing predictive biomarkers are imperfect. Here, we used an end-to-end deep learning algorithm to identify risk factors for lymph node metastasis (LNM) status in digitized histopathology slides of the primary CRC and its surrounding tissue. In two large population-based cohorts, we show that this system can predict the presence of more than one LNM in pT2 CRC patients with an area under the receiver operating curve (AUROC) of 0.733 (0.67–0.758) and patients with any LNM with an AUROC of 0.711 (0.597–0.797). Similarly, in pT1 CRC patients, the presence of more than one LNM or any LNM was predictable with an AUROC of 0.733 (0.644–0.778) and 0.567 (0.542–0.597), respectively. Based on these findings, we used the deep learning system to guide human pathology experts towards highly predictive regions for LNM in the whole slide images. This hybrid human observer and deep learning approach identified inflamed adipose tissue as the highest predictive feature for LNM presence. Our study is a first proof of concept that artificial intelligence (AI) systems may be able to discover potentially new biological mechanisms in cancer progression. Our deep learning algorithm is publicly available and can be used for biomarker discovery in any disease setting.
KW - AI
KW - artificial intelligence
KW - deep learning
KW - digital pathology
KW - early colorectal cancer
KW - inflamed adipose tissue
KW - metastasis
KW - new predictive biomarker
KW - prediction LNM
KW - pT1 and pT2 bowel cancer
U2 - 10.1002/path.5831
DO - 10.1002/path.5831
M3 - Journal article
C2 - 34738636
AN - SCOPUS:85121997885
VL - 256
SP - 269
EP - 281
JO - Journal of Pathology
JF - Journal of Pathology
SN - 0022-3417
IS - 3
ER -
ID: 305692536