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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

The quality and richness of feature maps extracted by convolution neuralnetworks (CNNs) and vision Transformers (ViTs) directly relate to the robustmodel performance. In medical computer vision, these information-rich featuresare crucial for detecting rare cases within large datasets. This work presentsthe "Scopeformer," a novel multi-CNN-ViT model for intracranial hemorrhageclassification in computed tomography (CT) images. The Scopeformer architectureis scalable and modular, which allows utilizing various CNN architectures asthe backbone with diversified output features and pre-training strategies. Wepropose effective feature projection methods to reduce redundancies amongCNN-generated features and to control the input size of ViTs. Extensiveexperiments with various Scopeformer models show that the model performance isproportional to the number of convolutional blocks employed in the featureextractor. Using multiple strategies, including diversifying the pre-trainingparadigms for CNNs, different pre-training datasets, and style transfertechniques, we demonstrate an overall improvement in the model performance atvarious computational budgets. Later, we propose smaller compute-efficientScopeformer versions with three different types of input and output ViTconfigurations. Efficient Scopeformers use four different pre-trained CNNarchitectures as feature extractors to increase feature richness. Our bestEfficient Scopeformer model achieved an accuracy of 96.94\% and a weightedlogarithmic loss of 0.083 with an eight times reduction in the number oftrainable parameters compared to the base Scopeformer. Another version of theEfficient Scopeformer model further reduced the parameter space by almost 17times with negligible performance reduction. Hybrid CNNs and ViTs might providethe desired feature richness for developing accurate medical computer visionmodels

Efficient Scopeformer: Towards Scalable and Rich Feature Extraction for Intracranial Hemorrhage Detection