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

Purpose: Automating tasks such as lung tumor localization and segmentation inradiological images can free valuable time for radiologists and other clinicalpersonnel. Convolutional neural networks may be suited for such tasks, butrequire substantial amounts of labeled data to train. Obtaining labeled data isa challenge, especially in the medical domain. Methods: This paper investigatesthe use of a teacher-student design to utilize datasets with different types ofsupervision to train an automatic model performing pulmonary tumor segmentationon computed tomography images. The framework consists of two models: thestudent that performs end-to-end automatic tumor segmentation and the teacherthat supplies the student additional pseudo-annotated data during training.Results: Using only a small proportion of semantically labeled data and a largenumber of bounding box annotated data, we achieved competitive performanceusing a teacher-student design. Models trained on larger amounts of semanticannotations did not perform better than those trained on teacher-annotateddata. Conclusions: Our results demonstrate the potential of utilizingteacher-student designs to reduce the annotation load, as less supervisedannotation schemes may be performed, without any real degradation insegmentation accuracy.

Teacher-Student Architecture for Mixed Supervised Lung Tumor Segmentation