Deep learning based classification of multi-label chest X-ray images via dual-weighted metric loss. (May 2023)
- Record Type:
- Journal Article
- Title:
- Deep learning based classification of multi-label chest X-ray images via dual-weighted metric loss. (May 2023)
- Main Title:
- Deep learning based classification of multi-label chest X-ray images via dual-weighted metric loss
- Authors:
- Jin, Yufei
Lu, Huijuan
Zhu, Wenjie
Huo, Wanli - Abstract:
- Abstract: —Thoracic disease, like many other diseases, can lead to complications. Existing multi-label medical image learning problems typically include rich pathological information, such as images, attributes, and labels, which are crucial for supplementary clinical diagnosis. However, the majority of contemporary efforts exclusively focus on regression from input to binary labels, ignoring the relationship between visual features and semantic vectors of labels. In addition, there is an imbalance in data amount between diseases, which frequently causes intelligent diagnostic systems to make erroneous disease predictions. Therefore, we aim to improve the accuracy of the multi-label classification of chest X-ray images. Chest X-ray14 pictures were utilized as the multi-label dataset for the experiments in this study. By fine-tuning the ConvNeXt network, we got visual vectors, which we combined with semantic vectors encoded by BioBert to map the two different forms of features into a common metric space and made semantic vectors the prototype of each class in metric space. The metric relationship between images and labels is then considered from the image level and disease category level, respectively, and a new dual-weighted metric loss function is proposed. Finally, the average AUC score achieved in the experiment reached 0.826, and our model outperformed the comparison models. Highlights: No need to extract each category feature from multi-labeled images as a prototype.Abstract: —Thoracic disease, like many other diseases, can lead to complications. Existing multi-label medical image learning problems typically include rich pathological information, such as images, attributes, and labels, which are crucial for supplementary clinical diagnosis. However, the majority of contemporary efforts exclusively focus on regression from input to binary labels, ignoring the relationship between visual features and semantic vectors of labels. In addition, there is an imbalance in data amount between diseases, which frequently causes intelligent diagnostic systems to make erroneous disease predictions. Therefore, we aim to improve the accuracy of the multi-label classification of chest X-ray images. Chest X-ray14 pictures were utilized as the multi-label dataset for the experiments in this study. By fine-tuning the ConvNeXt network, we got visual vectors, which we combined with semantic vectors encoded by BioBert to map the two different forms of features into a common metric space and made semantic vectors the prototype of each class in metric space. The metric relationship between images and labels is then considered from the image level and disease category level, respectively, and a new dual-weighted metric loss function is proposed. Finally, the average AUC score achieved in the experiment reached 0.826, and our model outperformed the comparison models. Highlights: No need to extract each category feature from multi-labeled images as a prototype. Proposed dual-weighted metric loss of image and label relationships. Combining image features with label semantic vectors can effectively improve multi-label classification performance. The effectiveness of the dual-weighted metric loss for multi-label classification is verified. … (more)
- Is Part Of:
- Computers in biology and medicine. Volume 157(2023)
- Journal:
- Computers in biology and medicine
- Issue:
- Volume 157(2023)
- Issue Display:
- Volume 157, Issue 2023 (2023)
- Year:
- 2023
- Volume:
- 157
- Issue:
- 2023
- Issue Sort Value:
- 2023-0157-2023-0000
- Page Start:
- Page End:
- Publication Date:
- 2023-05
- Subjects:
- Deep metric learning -- Multi-label learning -- ConvNeXt -- Chest X-ray
Medicine -- Data processing -- Periodicals
Biology -- Data processing -- Periodicals
610.285 - Journal URLs:
- http://www.sciencedirect.com/science/journal/00104825/ ↗
http://www.elsevier.com/journals ↗ - DOI:
- 10.1016/j.compbiomed.2023.106683 ↗
- Languages:
- English
- ISSNs:
- 0010-4825
- Deposit Type:
- Legaldeposit
- View Content:
- Available online (eLD content is only available in our Reading Rooms) ↗
- Physical Locations:
- British Library DSC - 3394.880000
British Library DSC - BLDSS-3PM
British Library HMNTS - ELD Digital store - Ingest File:
- 26775.xml