Scoring of tumor-infiltrating lymphocytes

From visual estimation to machine learning

on behalf of the International Immuno-Oncology Biomarker Working Group

Research output: Contribution to journalReview article

4 Citations (Scopus)

Abstract

The extent of tumor-infiltrating lymphocytes (TILs), along with immunomodulatory ligands, tumor-mutational burden and other biomarkers, has been demonstrated to be a marker of response to immune-checkpoint therapy in several cancers. Pathologists have therefore started to devise standardized visual approaches to quantify TILs for therapy prediction. However, despite successful standardization efforts visual TIL estimation is slow, with limited precision and lacks the ability to evaluate more complex properties such as TIL distribution patterns. Therefore, computational image analysis approaches are needed to provide standardized and efficient TIL quantification. Here, we discuss different automated TIL scoring approaches ranging from classical image segmentation, where cell boundaries are identified and the resulting objects classified according to shape properties, to machine learning-based approaches that directly classify cells without segmentation but rely on large amounts of training data. In contrast to conventional machine learning (ML) approaches that are often criticized for their “black-box” characteristics, we also discuss explainable machine learning. Such approaches render ML results interpretable and explain the computational decision-making process through high-resolution heatmaps that highlight TILs and cancer cells and therefore allow for quantification and plausibility checks in biomedical research and diagnostics.

Original languageEnglish (US)
Pages (from-to)151-157
Number of pages7
JournalSeminars in Cancer Biology
Volume52
DOIs
StatePublished - Oct 1 2018

Fingerprint

Tumor-Infiltrating Lymphocytes
Aptitude
Machine Learning
Tumor Burden
Biomedical Research
Neoplasms
Decision Making
Biomarkers
Ligands
Therapeutics

ASJC Scopus subject areas

  • Cancer Research

Cite this

Scoring of tumor-infiltrating lymphocytes : From visual estimation to machine learning. / on behalf of the International Immuno-Oncology Biomarker Working Group.

In: Seminars in Cancer Biology, Vol. 52, 01.10.2018, p. 151-157.

Research output: Contribution to journalReview article

on behalf of the International Immuno-Oncology Biomarker Working Group. / Scoring of tumor-infiltrating lymphocytes : From visual estimation to machine learning. In: Seminars in Cancer Biology. 2018 ; Vol. 52. pp. 151-157.
@article{90355ebc956b4265b33d8c1186e4914e,
title = "Scoring of tumor-infiltrating lymphocytes: From visual estimation to machine learning",
abstract = "The extent of tumor-infiltrating lymphocytes (TILs), along with immunomodulatory ligands, tumor-mutational burden and other biomarkers, has been demonstrated to be a marker of response to immune-checkpoint therapy in several cancers. Pathologists have therefore started to devise standardized visual approaches to quantify TILs for therapy prediction. However, despite successful standardization efforts visual TIL estimation is slow, with limited precision and lacks the ability to evaluate more complex properties such as TIL distribution patterns. Therefore, computational image analysis approaches are needed to provide standardized and efficient TIL quantification. Here, we discuss different automated TIL scoring approaches ranging from classical image segmentation, where cell boundaries are identified and the resulting objects classified according to shape properties, to machine learning-based approaches that directly classify cells without segmentation but rely on large amounts of training data. In contrast to conventional machine learning (ML) approaches that are often criticized for their “black-box” characteristics, we also discuss explainable machine learning. Such approaches render ML results interpretable and explain the computational decision-making process through high-resolution heatmaps that highlight TILs and cancer cells and therefore allow for quantification and plausibility checks in biomedical research and diagnostics.",
author = "{on behalf of the International Immuno-Oncology Biomarker Working Group} and F. Klauschen and M{\"u}ller, {K. R.} and A. Binder and M. Bockmayr and M. H{\"a}gele and P. Seegerer and S. Wienert and G. Pruneri and {de Maria}, S. and Sunil Badve and S. Michiels and Nielsen, {T. O.} and S. Adams and P. Savas and F. Symmans and S. Willis and T. Gruosso and M. Park and B. Haibe-Kains and B. Gallas and Thompson, {A. M.} and I. Cree and C. Sotiriou and C. Solinas and M. Preusser and Hewitt, {S. M.} and D. Rimm and G. Viale and S. Loi and S. Loibl and R. Salgado and C. Denkert",
year = "2018",
month = "10",
day = "1",
doi = "10.1016/j.semcancer.2018.07.001",
language = "English (US)",
volume = "52",
pages = "151--157",
journal = "Seminars in Cancer Biology",
issn = "1044-579X",
publisher = "Academic Press Inc.",

}

TY - JOUR

T1 - Scoring of tumor-infiltrating lymphocytes

T2 - From visual estimation to machine learning

AU - on behalf of the International Immuno-Oncology Biomarker Working Group

AU - Klauschen, F.

AU - Müller, K. R.

AU - Binder, A.

AU - Bockmayr, M.

AU - Hägele, M.

AU - Seegerer, P.

AU - Wienert, S.

AU - Pruneri, G.

AU - de Maria, S.

AU - Badve, Sunil

AU - Michiels, S.

AU - Nielsen, T. O.

AU - Adams, S.

AU - Savas, P.

AU - Symmans, F.

AU - Willis, S.

AU - Gruosso, T.

AU - Park, M.

AU - Haibe-Kains, B.

AU - Gallas, B.

AU - Thompson, A. M.

AU - Cree, I.

AU - Sotiriou, C.

AU - Solinas, C.

AU - Preusser, M.

AU - Hewitt, S. M.

AU - Rimm, D.

AU - Viale, G.

AU - Loi, S.

AU - Loibl, S.

AU - Salgado, R.

AU - Denkert, C.

PY - 2018/10/1

Y1 - 2018/10/1

N2 - The extent of tumor-infiltrating lymphocytes (TILs), along with immunomodulatory ligands, tumor-mutational burden and other biomarkers, has been demonstrated to be a marker of response to immune-checkpoint therapy in several cancers. Pathologists have therefore started to devise standardized visual approaches to quantify TILs for therapy prediction. However, despite successful standardization efforts visual TIL estimation is slow, with limited precision and lacks the ability to evaluate more complex properties such as TIL distribution patterns. Therefore, computational image analysis approaches are needed to provide standardized and efficient TIL quantification. Here, we discuss different automated TIL scoring approaches ranging from classical image segmentation, where cell boundaries are identified and the resulting objects classified according to shape properties, to machine learning-based approaches that directly classify cells without segmentation but rely on large amounts of training data. In contrast to conventional machine learning (ML) approaches that are often criticized for their “black-box” characteristics, we also discuss explainable machine learning. Such approaches render ML results interpretable and explain the computational decision-making process through high-resolution heatmaps that highlight TILs and cancer cells and therefore allow for quantification and plausibility checks in biomedical research and diagnostics.

AB - The extent of tumor-infiltrating lymphocytes (TILs), along with immunomodulatory ligands, tumor-mutational burden and other biomarkers, has been demonstrated to be a marker of response to immune-checkpoint therapy in several cancers. Pathologists have therefore started to devise standardized visual approaches to quantify TILs for therapy prediction. However, despite successful standardization efforts visual TIL estimation is slow, with limited precision and lacks the ability to evaluate more complex properties such as TIL distribution patterns. Therefore, computational image analysis approaches are needed to provide standardized and efficient TIL quantification. Here, we discuss different automated TIL scoring approaches ranging from classical image segmentation, where cell boundaries are identified and the resulting objects classified according to shape properties, to machine learning-based approaches that directly classify cells without segmentation but rely on large amounts of training data. In contrast to conventional machine learning (ML) approaches that are often criticized for their “black-box” characteristics, we also discuss explainable machine learning. Such approaches render ML results interpretable and explain the computational decision-making process through high-resolution heatmaps that highlight TILs and cancer cells and therefore allow for quantification and plausibility checks in biomedical research and diagnostics.

UR - http://www.scopus.com/inward/record.url?scp=85049578294&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85049578294&partnerID=8YFLogxK

U2 - 10.1016/j.semcancer.2018.07.001

DO - 10.1016/j.semcancer.2018.07.001

M3 - Review article

VL - 52

SP - 151

EP - 157

JO - Seminars in Cancer Biology

JF - Seminars in Cancer Biology

SN - 1044-579X

ER -