Publications by authors named "Tongxu Lin"
Multi-Label Chest X-Ray Image Classification with Single Positive Labels.
IEEE Trans Med Imaging
July 2024
Deep learning approaches for multi-label Chest X-ray (CXR) images classification usually require large-scale datasets. However, acquiring such datasets with full annotations is costly, time-consuming, and prone to noisy labels. Therefore, we introduce a weakly supervised learning problem called Single Positive Multi-label Learning (SPML) into CXR images classification (abbreviated as SPML-CXR), in which only one positive label is annotated per image.
View Article and Find Full Text PDFArticle Synopsis
- MRI-based multi-modal brain tumor segmentation (MBTS) has gained interest due to the effectiveness of non-invasive imaging, but existing studies often struggle with limited data collection.
- The authors introduce a novel quaternion mutual learning strategy (QMLS) that includes a voxel-wise lesion knowledge mutual learning mechanism and a quaternion multi-modal feature learning module, enhancing the model's ability to learn from sparse data.
- QMLS significantly outperforms current methods in terms of performance and computational efficiency, making it a promising advancement for automatic brain tumor segmentation in clinical settings.
Article Synopsis
- Breast tumor segmentation is challenging due to the tumors' irregular shapes, but deep convolution networks have shown promise in improving segmentation results.
- To enhance the segmentation performance, a novel Shape-Guided Segmentation (SGS) framework is proposed, incorporating a Shape Guiding Block (SGB) and a Shared Classification Layer (SCL) to retain and utilize shape information across samples.
- Experiments indicate the SGS framework outperforms other advanced segmentation methods on both private and public datasets, with the source code available on GitHub for further use and study.