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

Convolutional neural networks (CNNs) achieved the state-of-the-artperformance in medical image segmentation due to their ability to extracthighly complex feature representations. However, it is argued in recent studiesthat traditional CNNs lack the intelligence to capture long-term dependenciesof different image regions. Following the success of applying Transformermodels on natural language processing tasks, the medical image segmentationfield has also witnessed growing interest in utilizing Transformers, due totheir ability to capture long-range contextual information. However, unlikeCNNs, Transformers lack the ability to learn local feature representations.Thus, to fully utilize the advantages of both CNNs and Transformers, we proposea hybrid encoder-decoder segmentation model (ConvTransSeg). It consists of amulti-layer CNN as the encoder for feature learning and the correspondingmulti-level Transformer as the decoder for segmentation prediction. The encoderand decoder are interconnected in a multi-resolution manner. We compared ourmethod with many other state-of-the-art hybrid CNN and Transformer segmentationmodels on binary and multiple class image segmentation tasks using severalpublic medical image datasets, including skin lesion, polyp, cell and braintissue. The experimental results show that our method achieves overall the bestperformance in terms of Dice coefficient and average symmetric surface distancemeasures with low model complexity and memory consumption. In contrast to mostTransformer-based methods that we compared, our method does not require the useof pre-trained models to achieve similar or better performance. The code isfreely available for research purposes on Github: (the link will be added uponacceptance).

ConvTransSeg: A Multi-resolution Convolution-Transformer Network for Medical Image Segmentation