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

In medical image analysis, automated segmentation of multi-componentanatomical structures, which often have a spectrum of potential anomalies andpathologies, is a challenging task. In this work, we develop a multi-stepapproach using U-Net-based neural networks to initially detect anomalies (bonemarrow lesions, bone cysts) in the distal femur, proximal tibia and patellafrom 3D magnetic resonance (MR) images of the knee in individuals with varyinggrades of osteoarthritis. Subsequently, the extracted data are used fordownstream tasks involving semantic segmentation of individual bone andcartilage volumes as well as bone anomalies. For anomaly detection, theU-Net-based models were developed to reconstruct the bone profiles of the femurand tibia in images via inpainting so anomalous bone regions could be replacedwith close to normal appearances. The reconstruction error was used to detectbone anomalies. A second anomaly-aware network, which was compared toanomaly-na\"ive segmentation networks, was used to provide a final automatedsegmentation of the femoral, tibial and patellar bones and cartilages from theknee MR images containing a spectrum of bone anomalies. The anomaly-awaresegmentation approach provided up to 58% reduction in Hausdorff distances forbone segmentations compared to the results from the anomaly-na\"ivesegmentation networks. In addition, the anomaly-aware networks were able todetect bone lesions in the MR images with greater sensitivity and specificity(area under the receiver operating characteristic curve [AUC] up to 0.896)compared to the anomaly-na\"ive segmentation networks (AUC up to 0.874).

Automated anomaly-aware 3D segmentation of bones and cartilages in knee MR images from the Osteoarthritis Initiative