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/

Zendy Plus

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

Zendy Tools

Zendy Open

Accurate and automated tumor segmentation is highly desired since it has thegreat potential to increase the efficiency and reproducibility of computingmore complete tumor measurements and imaging biomarkers, comparing to (oftenpartial) human measurements. This is probably the only viable means to enablethe large-scale clinical oncology patient studies that utilize medical imaging.Deep learning approaches have shown robust segmentation performances forcertain types of tumors, e.g., brain tumors in MRI imaging, when a trainingdataset with plenty of pixel-level fully-annotated tumor images is available.However, more than often, we are facing the challenge that only (very) limitedannotations are feasible to acquire, especially for hard tumors. Pancreaticductal adenocarcinoma (PDAC) segmentation is one of the most challenging tumorsegmentation tasks, yet critically important for clinical needs. Previous workon PDAC segmentation is limited to the moderate amounts of annotated patientimages (n<300) from venous or venous+arterial phase CT scans. Based on a newself-learning framework, we propose to train the PDAC segmentation model usinga much larger quantity of patients (n~=1,000), with a mix of annotated andun-annotated venous or multi-phase CT images. Pseudo annotations are generatedby combining two teacher models with different PDAC segmentation specialties onunannotated images, and can be further refined by a teaching assistant modelthat identifies associated vessels around the pancreas. A student model istrained on both manual and pseudo annotated multi-phase images. Experimentresults show that our proposed method provides an absolute improvement of 6.3%Dice score over the strong baseline of nnUNet trained on annotated images,achieving the performance (Dice = 0.71) similar to the inter-observervariability between radiologists.

Robust Pancreatic Ductal Adenocarcinoma Segmentation with Multi-Institutional Multi-Phase Partially-Annotated CT Scans