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

Robust automated organ segmentation is a prerequisite for computer-aideddiagnosis (CAD), quantitative imaging analysis and surgical assistance. Forhigh-variability organs such as the pancreas, previous approaches reportundesirably low accuracies. We present a bottom-up approach for pancreassegmentation in abdominal CT scans that is based on a hierarchy of informationpropagation by classifying image patches at different resolutions; andcascading superpixels. There are four stages: 1) decomposing CT slice images asa set of disjoint boundary-preserving superpixels; 2) computing pancreas classprobability maps via dense patch labeling; 3) classifying superpixels bypooling both intensity and probability features to form empirical statistics incascaded random forest frameworks; and 4) simple connectivity basedpost-processing. The dense image patch labeling are conducted by: efficientrandom forest classifier on image histogram, location and texture features; andmore expensive (but with better specificity) deep convolutional neural networkclassification on larger image windows (with more spatial contexts). Evaluationof the approach is performed on a database of 80 manually segmented CT volumesin six-fold cross-validation (CV). Our achieved results are comparable, orbetter than the state-of-the-art methods (evaluated by"leave-one-patient-out"), with Dice 70.7% and Jaccard 57.9%. The computationalefficiency has been drastically improved in the order of 6~8 minutes, comparingwith others of ~10 hours per case. Finally, we implement a multi-atlas labelfusion (MALF) approach for pancreas segmentation using the same datasets. Undersix-fold CV, our bottom-up segmentation method significantly outperforms itsMALF counterpart: (70.7 +/- 13.0%) versus (52.5 +/- 20.8%) in Dice. Deep CNNpatch labeling confidences offer more numerical stability, reflected by smallerstandard deviations.

A Bottom-up Approach for Pancreas Segmentation using Cascaded Superpixels and (Deep) Image Patch Labeling