Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

LMFLOSS: A Hybrid Loss For Imbalanced Medical Image Classification

Abu Adnan Sadi, Labib Chowdhury, Nusrat Jahan, Mohammad Newaz Sharif Rafi, Radeya Chowdhury, Faisal Ahamed Khan, Nabeel Mohammed

TL;DR

This study proposes a novel framework called Large Margin aware Focal (LMF) loss to mitigate the class imbalance problem in medical imaging and demonstrates a simple and effective approach to addressing the class imbalance problem in medical imaging datasets.

Abstract

With advances in digital technology, the classification of medical images has become a crucial step for image-based clinical decision support systems. Automatic medical image classification represents a pivotal domain where the use of AI holds the potential to create a significant social impact. However, several challenges act as obstacles to the development of practical and effective solutions. One of these challenges is the prevalent class imbalance problem in most medical imaging datasets. As a result, existing AI techniques, particularly deep-learning-based methodologies, often underperform in such scenarios. In this study, we propose a novel framework called Large Margin aware Focal (LMF) loss to mitigate the class imbalance problem in medical imaging. The LMF loss represents a linear combination of two loss functions optimized by two hyperparameters. This framework harnesses the distinct characteristics of both loss functions by enforcing wider margins for minority classes while simultaneously emphasizing challenging samples found in the datasets. We perform rigorous experiments on three neural network architectures and with four medical imaging datasets. We provide empirical evidence that our proposed framework consistently outperforms other baseline methods, showing an improvement of 2%-9% in macro-f1 scores. Through class-wise analysis of f1 scores, we also demonstrate how the proposed framework can significantly improve performance for minority classes. The results of our experiments show that our proposed framework can perform consistently well across different architectures and datasets. Overall, our study demonstrates a simple and effective approach to addressing the class imbalance problem in medical imaging datasets. We hope our work will inspire new research toward a more generalized approach to medical image classification.

LMFLOSS: A Hybrid Loss For Imbalanced Medical Image Classification

TL;DR

This study proposes a novel framework called Large Margin aware Focal (LMF) loss to mitigate the class imbalance problem in medical imaging and demonstrates a simple and effective approach to addressing the class imbalance problem in medical imaging datasets.

Abstract

With advances in digital technology, the classification of medical images has become a crucial step for image-based clinical decision support systems. Automatic medical image classification represents a pivotal domain where the use of AI holds the potential to create a significant social impact. However, several challenges act as obstacles to the development of practical and effective solutions. One of these challenges is the prevalent class imbalance problem in most medical imaging datasets. As a result, existing AI techniques, particularly deep-learning-based methodologies, often underperform in such scenarios. In this study, we propose a novel framework called Large Margin aware Focal (LMF) loss to mitigate the class imbalance problem in medical imaging. The LMF loss represents a linear combination of two loss functions optimized by two hyperparameters. This framework harnesses the distinct characteristics of both loss functions by enforcing wider margins for minority classes while simultaneously emphasizing challenging samples found in the datasets. We perform rigorous experiments on three neural network architectures and with four medical imaging datasets. We provide empirical evidence that our proposed framework consistently outperforms other baseline methods, showing an improvement of 2%-9% in macro-f1 scores. Through class-wise analysis of f1 scores, we also demonstrate how the proposed framework can significantly improve performance for minority classes. The results of our experiments show that our proposed framework can perform consistently well across different architectures and datasets. Overall, our study demonstrates a simple and effective approach to addressing the class imbalance problem in medical imaging datasets. We hope our work will inspire new research toward a more generalized approach to medical image classification.
Paper Structure (21 sections, 8 equations, 4 figures, 6 tables)

This paper contains 21 sections, 8 equations, 4 figures, 6 tables.

Table of Contents

  1. Abstract
  2. Introduction
  3. Materials and methods
  4. Results and discussion
  5. Conclusion
  6. Acknowledgments
  7. Data availability statement

Figures (4)

  • Figure 1: Macro-f1 scores of the proposed method and compared to four other pre-existing techniques for four different medical image datasets. Each error bar depicts the mean of the macro-f1 scores obtained from three different network architectures, along with its average deviation. The proposed LMF-loss achieves higher average macro-f1 scores for all four datasets.
  • Figure 2: Per-class image distribution of all four datasets. (a) ODIR-5K, (b) HAM-10K, (c) ISIC-2019, and (d) COVID-19 Radiography.
  • Figure 3: Grad-CAM attention map visualization and comparison on test samples from all four datasets. Green bounding boxes depict annotations obtained from a doctor.
  • Figure 4: Mean macro-f1 scores obtained from each hyperparameter setting, along with their average deviation. We obtained the highest average macro-f1 score when we only modified the value of the LMF-loss, and was set to 1.0.