An Ordinal Regression Framework for a Deep Learning Based Severity Assessment for Chest Radiographs

arXiv (Cornell University)

Abstract

This study investigates the application of ordinal regression methods for categorizing disease severity in chest radiographs. We propose a framework that divides the ordinal regression problem into three parts: a model, a target function, and a classification function. Different encoding methods, including one-hot, Gaussian, progress-bar, and our soft-progress-bar, are applied using ResNet50 and ViT-B-16 deep learning models. We show that the choice of encoding has a strong impact on performance and that the best encoding depends on the chosen weighting of Cohen's kappa and also on the model architecture used. We make our code publicly available on GitHub.

Authors 8

1. Patrick Wienholt Aachen Department of Diagnostic and Interventional Radiology

   RWTH Aachen University

   Affiliation as printed

   Department of Diagnostic and Interventional Radiology , RWTH Aachen University Hospital , Aachen , Germany

2. Alexander Hermans Aachen Department of Diagnostic and Interventional Radiology Visual Computing Institute

   RWTH Aachen University

   Affiliation as printed

   Department of Diagnostic and Interventional Radiology , RWTH Aachen University Hospital , Aachen , Germany

   Visual Computing Institute , RWTH Aachen University , Aachen , Germany

3. Firas Khader Aachen Department of Diagnostic and Interventional Radiology

   RWTH Aachen University

   Affiliation as printed

   Department of Diagnostic and Interventional Radiology , RWTH Aachen University Hospital , Aachen , Germany

4. Behrus Puladi Aachen Department of Oral and Maxillofacial Surgery Institute of Medical Informatics

   RWTH Aachen University

   Affiliation as printed

   Department of Oral and Maxillofacial Surgery , RWTH Aachen University Hospital , Aachen , Germany

   Institute of Medical Informatics , RWTH Aachen University Hospital , Aachen , Germany

5. Bastian Leibe Aachen Visual Computing Institute

   RWTH Aachen University

   Affiliation as printed

   Visual Computing Institute , RWTH Aachen University , Aachen , Germany

6. Christiane Katharina Kuhl Aachen Department of Diagnostic and Interventional Radiology

   RWTH Aachen University

   Affiliation as printed

   Department of Diagnostic and Interventional Radiology , RWTH Aachen University Hospital , Aachen , Germany

7. Sven Nebelung Aachen Department of Diagnostic and Interventional Radiology

   RWTH Aachen University

   Affiliation as printed

   Department of Diagnostic and Interventional Radiology , RWTH Aachen University Hospital , Aachen , Germany

8. Daniel Truhn Aachen Department of Diagnostic and Interventional Radiology

   RWTH Aachen University

   Affiliation as printed

   Department of Diagnostic and Interventional Radiology , RWTH Aachen University Hospital , Aachen , Germany