[135647] |
Title: Skin Lesion Classification Using CNNs With Patch-Based Attention and Diagnosis-Guided Loss Weighting. |
Written by: N. Gessert and T. Sentker and F. Madesta and R. Schmitz and H. Kniep and I. Baltruschat and R. Werner and A. Schlaefer |
in: <em>IEEE Transactions on Biomedical Engineering</em>. Feb (2020). |
Volume: <strong>67</strong>. Number: (2), |
on pages: 495-503 |
Chapter: |
Editor: |
Publisher: |
Series: |
Address: |
Edition: |
ISBN: |
how published: |
Organization: |
School: |
Institution: |
Type: |
DOI: 10.1109/TBME.2019.2915839 |
URL: https://arxiv.org/abs/1905.02793 |
ARXIVID: |
PMID: |
Note:
Abstract: Objective: This paper addresses two key problems of skin lesion classification. The first problem is the effective use of high-resolution images with pretrained standard architectures for image classification. The second problem is the high-class imbalance encountered in real-world multi-class datasets. Methods: To use high-resolution images, we propose a novel patch-based attention architecture that provides global context between small, high-resolution patches. We modify three pretrained architectures and study the performance of patch-based attention. To counter class imbalance problems, we compare oversampling, balanced batch sampling, and class-specific loss weighting. Additionally, we propose a novel diagnosis-guided loss weighting method that takes the method used for ground-truth annotation into account. Results: Our patch-based attention mechanism outperforms previous methods and improves the mean sensitivity by 7% . Class balancing significantly improves the mean sensitivity and we show that our diagnosis-guided loss weighting method improves the mean sensitivity by 3% over normal loss balancing. Conclusion: The novel patch-based attention mechanism can be integrated into pretrained architectures and provides global context between local patches while outperforming other patch-based methods. Hence, pretrained architectures can be readily used with high-resolution images without downsampling. The new diagnosis-guided loss weighting method outperforms other methods and allows for effective training when facing class imbalance. Significance: The proposed methods improve automatic skin lesion classification. They can be extended to other clinical applications where high-resolution image data and class imbalance are relevant