English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Current state-of-the-art medical image segmentation methods prioritizeaccuracy but often at the expense of increased computational demands and largermodel sizes. Applying these large-scale models to the relatively limited scaleof medical image datasets tends to induce redundant computation, complicatingthe process without the necessary benefits. This approach not only addscomplexity but also presents challenges for the integration and deployment oflightweight models on edge devices. For instance, recent transformer-basedmodels have excelled in 2D and 3D medical image segmentation due to theirextensive receptive fields and high parameter count. However, theireffectiveness comes with a risk of overfitting when applied to small datasetsand often neglects the vital inductive biases of Convolutional Neural Networks(CNNs), essential for local feature representation. In this work, we proposePMFSNet, a novel medical imaging segmentation model that effectively balancesglobal and local feature processing while avoiding the computational redundancytypical in larger models. PMFSNet streamlines the UNet-based hierarchicalstructure and simplifies the self-attention mechanism's computationalcomplexity, making it suitable for lightweight applications. It incorporates aplug-and-play PMFS block, a multi-scale feature enhancement module based onattention mechanisms, to capture long-term dependencies. Extensivecomprehensive results demonstrate that even with a model (less than 1 millionparameters), our method achieves superior performance in various segmentationtasks across different data scales. It achieves (IoU) metrics of 84.68%,82.02%, and 78.82% on public datasets of teeth CT (CBCT), ovarian tumorsultrasound(MMOTU), and skin lesions dermoscopy images (ISIC 2018),respectively. The source code is available athttps://github.com/yykzjh/PMFSNet.

PMFSNet: Polarized Multi-scale Feature Self-attention Network For Lightweight Medical Image Segmentation