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

Deep convolutional neural networks (CNNs) have been widely used for medicalimage segmentation. In most studies, only the output layer is exploited tocompute the final segmentation results and the hidden representations of thedeep learned features have not been well understood. In this paper, we proposea prototype segmentation (ProtoSeg) method to compute a binary segmentation mapbased on deep features. We measure the segmentation abilities of the featuresby computing the Dice between the feature segmentation map and ground-truth,named as the segmentation ability score (SA score for short). The correspondingSA score can quantify the segmentation abilities of deep features in differentlayers and units to understand the deep neural networks for segmentation. Inaddition, our method can provide a mean SA score which can give a performanceestimation of the output on the test images without ground-truth. Finally, weuse the proposed ProtoSeg method to compute the segmentation map directly oninput images to further understand the segmentation ability of each inputimage. Results are presented on segmenting tumors in brain MRI, lesions in skinimages, COVID-related abnormality in CT images, prostate segmentation inabdominal MRI, and pancreatic mass segmentation in CT images. Our method canprovide new insights for interpreting and explainable AI systems for medicalimage segmentation.  Our code is available on: \url{https://github.com/shengfly/ProtoSeg}.

Segmentation Ability Map: Interpret deep features for medical image segmentation