Lightweight Transformer Backbone for Medical Object Detection
This addresses the problem of label scarcity in medical imaging for clinicians, though it is incremental as it modifies existing Transformer methods for a specific domain.
The paper tackles lesion detection in digital breast tomosynthesis by proposing a lightweight Transformer backbone that connects tumor feature patches with healthy backgrounds, improving accuracy and reducing labeled data requirements. Experiments show it outperforms Faster R-CNN and SWIN transformer models with augmented data.
Lesion detection in digital breast tomosynthesis (DBT) is an important and a challenging problem characterized by a low prevalence of images containing tumors. Due to the label scarcity problem, large deep learning models and computationally intensive algorithms are likely to fail when applied to this task. In this paper, we present a practical yet lightweight backbone to improve the accuracy of tumor detection. Specifically, we propose a novel modification of visual transformer (ViT) on image feature patches to connect the feature patches of a tumor with healthy backgrounds of breast images and form a more robust backbone for tumor detection. To the best of our knowledge, our model is the first work of Transformer backbone object detection for medical imaging. Our experiments show that this model can considerably improve the accuracy of lesion detection and reduce the amount of labeled data required in typical ViT. We further show that with additional augmented tumor data, our model significantly outperforms the Faster R-CNN model and state-of-the-art SWIN transformer model.