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

Breast cancer diagnosis challenges both patients and clinicians, with earlydetection being crucial for effective treatment. Ultrasound imaging plays a keyrole in this, but its utility is hampered by the need for precise lesionsegmentation-a task that is both time-consuming and labor-intensive. To addressthese challenges, we propose a new framework: a morphology-enhanced, ClassActivation Map (CAM)-guided model, which is optimized using a computer visionfoundation model known as SAM. This innovative framework is specificallydesigned for weakly supervised lesion segmentation in early-stage breastultrasound images. Our approach uniquely leverages image-level annotations,which removes the requirement for detailed pixel-level annotation. Initially,we perform a preliminary segmentation using breast lesion morphology knowledge.Following this, we accurately localize lesions by extracting semanticinformation through a CAM-based heatmap. These two elements are then fusedtogether, serving as a prompt to guide the SAM in performing refinedsegmentation. Subsequently, post-processing techniques are employed to rectifytopological errors made by the SAM. Our method not only simplifies thesegmentation process but also attains accuracy comparable to supervisedlearning methods that rely on pixel-level annotation. Our framework achieves aDice score of 74.39% on the test set, demonstrating compareable performancewith supervised learning methods. Additionally, it outperforms a supervisedlearning model, in terms of the Hausdorff distance, scoring 24.27 compared toDeeplabv3+'s 32.22. These experimental results showcase its feasibility andsuperior performance in integrating weakly supervised learning with SAM. Thecode is made available at: https://github.com/YueXin18/MorSeg-CAM-SAM.

Morphology-Enhanced CAM-Guided SAM for weakly supervised Breast Lesion Segmentation