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

We propose a fast beam orientation selection method, based on deep neuralnetworks (DNN), capable of developing a plan comparable to those by thestate-of-the-art column generation method. The novelty of Our model lies in itssupervised learning structure, the DNN architecture, and ability to learn fromanatomical features to predict dosimetrically suitable beam orientationswithout using the dosimetric information from the candidate beams, a timeconsuming and computationally expensive process. This may save hours ofcomputation. A supervised DNN is trained to mimic the column generationalgorithm, which iteratively chooses beam orientations by calculating beamfitness values based on the KKT optimality conditions. The dataset contains 70prostate cancer patients. The DNN trained over 400 epochs, each with 2500steps, using the Adam optimizer and a 6-fold cross-validation technique. Theaverage and standard deviation of training, validation, and testing lossfunctions among the 6-folds were at most 1.44%. The differences in the dosecoverage of PTV between plans generated by column generation and by DNN were0.2%. The average dose differences received by organs at risk were between 1and 6 percent: bladder had the smallest average difference, then rectum, leftand right femoral heads. The dose received by body had an average difference of0.1%. In the training phase of the proposed method, the model learns thesuitable beam orientations based on the anatomical features of patients andomits time intensive calculations of dose influence matrices for all possiblecandidate beams. Solving the Fluence Map Optimization to get the finaltreatment plan requires calculating dose influence matrices only for theselected beams. The proposed DNN is a fast beam orientation selection methodbased that selects beam orientations in seconds and is therefore suitable forclinical routines.

A Fast Deep Learning Approach for Beam Orientation Optimization for Prostate Cancer IMRT Treatments