AUC margin loss for limited, imbalanced and noisy medical image diagnosis: a case study on CheXpert5000

Research output: Contribution to journalArticleResearchpeer review

View graph of relations

Details

Original languageEnglish
Pages (from-to)658-661
Number of pages4
JournalCurrent Directions in Biomedical Engineering
Volume9
Issue number1
Publication statusPublished - 1 Sept 2023

Abstract

The AUC margin loss is a valuable loss function for medical image classification as it addresses the problems of imbalanced and noisy labels. It is used by the current winner of the CheXpert competition. The CheXpert dataset is a large dataset (200k+ images), however datasets in the range of 1k-10k medical datasets are much more common. This raises the question if optimizing AUC margin loss also is effective in scenarios with limited data.We compare AUC margin loss optimization to binary cross-entropy on limited, imbalanced and noisy CheXpert5000, a subset of CheXpert dataset. We show that AUC margin loss is beneficial for limited data and considerably improves accuracy in the presence of label noise. It also improves out-of-box calibration.

Keywords

    CheXpert, Computer-Aided-Diagnosis, Deep Learning, Label Imbalance, Noisy Labels

ASJC Scopus subject areas

Cite this

AUC margin loss for limited, imbalanced and noisy medical image diagnosis: a case study on CheXpert5000. / Ihler, Sontje; Kuhnke, Felix.
In: Current Directions in Biomedical Engineering, Vol. 9, No. 1, 01.09.2023, p. 658-661.

Research output: Contribution to journalArticleResearchpeer review

Ihler S, Kuhnke F. AUC margin loss for limited, imbalanced and noisy medical image diagnosis: a case study on CheXpert5000. Current Directions in Biomedical Engineering. 2023 Sept 1;9(1):658-661. doi: 10.1515/cdbme-2023-1165
Ihler, Sontje ; Kuhnke, Felix. / AUC margin loss for limited, imbalanced and noisy medical image diagnosis : a case study on CheXpert5000. In: Current Directions in Biomedical Engineering. 2023 ; Vol. 9, No. 1. pp. 658-661.
Download
@article{8e225bf3364f41aeae5126c6b3b2040a,
title = "AUC margin loss for limited, imbalanced and noisy medical image diagnosis: a case study on CheXpert5000",
abstract = "The AUC margin loss is a valuable loss function for medical image classification as it addresses the problems of imbalanced and noisy labels. It is used by the current winner of the CheXpert competition. The CheXpert dataset is a large dataset (200k+ images), however datasets in the range of 1k-10k medical datasets are much more common. This raises the question if optimizing AUC margin loss also is effective in scenarios with limited data.We compare AUC margin loss optimization to binary cross-entropy on limited, imbalanced and noisy CheXpert5000, a subset of CheXpert dataset. We show that AUC margin loss is beneficial for limited data and considerably improves accuracy in the presence of label noise. It also improves out-of-box calibration.",
keywords = "CheXpert, Computer-Aided-Diagnosis, Deep Learning, Label Imbalance, Noisy Labels",
author = "Sontje Ihler and Felix Kuhnke",
year = "2023",
month = sep,
day = "1",
doi = "10.1515/cdbme-2023-1165",
language = "English",
volume = "9",
pages = "658--661",
number = "1",

}

Download

TY - JOUR

T1 - AUC margin loss for limited, imbalanced and noisy medical image diagnosis

T2 - a case study on CheXpert5000

AU - Ihler, Sontje

AU - Kuhnke, Felix

PY - 2023/9/1

Y1 - 2023/9/1

N2 - The AUC margin loss is a valuable loss function for medical image classification as it addresses the problems of imbalanced and noisy labels. It is used by the current winner of the CheXpert competition. The CheXpert dataset is a large dataset (200k+ images), however datasets in the range of 1k-10k medical datasets are much more common. This raises the question if optimizing AUC margin loss also is effective in scenarios with limited data.We compare AUC margin loss optimization to binary cross-entropy on limited, imbalanced and noisy CheXpert5000, a subset of CheXpert dataset. We show that AUC margin loss is beneficial for limited data and considerably improves accuracy in the presence of label noise. It also improves out-of-box calibration.

AB - The AUC margin loss is a valuable loss function for medical image classification as it addresses the problems of imbalanced and noisy labels. It is used by the current winner of the CheXpert competition. The CheXpert dataset is a large dataset (200k+ images), however datasets in the range of 1k-10k medical datasets are much more common. This raises the question if optimizing AUC margin loss also is effective in scenarios with limited data.We compare AUC margin loss optimization to binary cross-entropy on limited, imbalanced and noisy CheXpert5000, a subset of CheXpert dataset. We show that AUC margin loss is beneficial for limited data and considerably improves accuracy in the presence of label noise. It also improves out-of-box calibration.

KW - CheXpert

KW - Computer-Aided-Diagnosis

KW - Deep Learning

KW - Label Imbalance

KW - Noisy Labels

UR - http://www.scopus.com/inward/record.url?scp=85173232202&partnerID=8YFLogxK

U2 - 10.1515/cdbme-2023-1165

DO - 10.1515/cdbme-2023-1165

M3 - Article

AN - SCOPUS:85173232202

VL - 9

SP - 658

EP - 661

JO - Current Directions in Biomedical Engineering

JF - Current Directions in Biomedical Engineering

IS - 1

ER -

By the same author(s)