Exploiting Scale-Variant Attention for Segmenting Small Medical Objects

2024-07-10

AI Summary

To address the challenge of segmenting sub-pixel-scale lesions (<1% of image area) in medical imaging, this paper proposes the Scale-Variant Attention Network (SvANet). Methodologically, SvANet innovatively integrates a scale-variant attention mechanism, a cross-scale guidance module, and Monte Carlo attention, while embedding a vision transformer to mitigate feature distortion caused by deep CNN compression. It adopts a U-Net–style encoder-decoder architecture to enable multi-granularity feature collaboration. Notably, this work is the first to introduce Monte Carlo attention into small-object segmentation, significantly enhancing robustness in modeling fine anatomical structures. The synergistic design of scale-variant attention and cross-scale guidance effectively suppresses inherent CNN artifacts. Evaluated on seven public benchmarks, SvANet achieves Dice scores of 72.58%–96.12%, consistently outperforming state-of-the-art methods—particularly excelling in ultra-small lesion segmentation tasks such as renal tumors and skin lesions.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Multi-instance/Multi-view Learning
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies

Abstract

Early detection and accurate diagnosis can predict the risk of malignant disease transformation, thereby increasing the probability of effective treatment. Identifying mild syndrome with small pathological regions serves as an ominous warning and is fundamental in the early diagnosis of diseases. While deep learning algorithms, particularly convolutional neural networks (CNNs), have shown promise in segmenting medical objects, analyzing small areas in medical images remains challenging. This difficulty arises due to information losses and compression defects from convolution and pooling operations in CNNs, which become more pronounced as the network deepens, especially for small medical objects. To address these challenges, we propose a novel scale-variant attention-based network (SvANet) for accurately segmenting small-scale objects in medical images. The SvANet consists of scale-variant attention, cross-scale guidance, Monte Carlo attention, and vision transformer, which incorporates cross-scale features and alleviates compression artifacts for enhancing the discrimination of small medical objects. Quantitative experimental results demonstrate the superior performance of SvANet, achieving 96.12%, 96.11%, 89.79%, 84.15%, 80.25%, 73.05%, and 72.58% in mean Dice coefficient for segmenting kidney tumors, skin lesions, hepatic tumors, polyps, surgical excision cells, retinal vasculatures, and sperms, which occupy less than 1% of the image areas in KiTS23, ISIC 2018, ATLAS, PolypGen, TissueNet, FIVES, and SpermHealth datasets, respectively.

Problem

• Segmenting small medical objects accurately in images
• Addressing information loss in CNNs for tiny regions
• Improving early disease diagnosis via precise detection

Innovation

• Scale-variant attention for small object segmentation
• Cross-scale guidance to enhance feature discrimination
• Monte Carlo attention with vision transformer integration