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 present a hybrid algorithm to estimate lung nodule malignancy thatcombines imaging biomarkers from Radiologist's annotation with imageclassification of CT scans. Our algorithm employs a 3D Convolutional NeuralNetwork (CNN) as well as a Random Forest in order to combine CT imagery withbiomarker annotation and volumetric radiomic features. We analyze and comparethe performance of the algorithm using only imagery, only biomarkers, combinedimagery + biomarkers, combined imagery + volumetric radiomic features andfinally the combination of imagery + biomarkers + volumetric features in orderto classify the suspicion level of nodule malignancy. The National CancerInstitute (NCI) Lung Image Database Consortium (LIDC) IDRI dataset is used totrain and evaluate the classification task. We show that the incorporation ofsemi-supervised learning by means of K-Nearest-Neighbors (KNN) can increase theavailable training sample size of the LIDC-IDRI thereby further improving theaccuracy of malignancy estimation of most of the models tested although thereis no significant improvement with the use of KNN semi-supervised learning ifimage classification with CNNs and volumetric features are combined withdescriptive biomarkers. Unexpectedly, we also show that a model using imagebiomarkers alone is more accurate than one that combines biomarkers withvolumetric radiomics, 3D CNNs, and semi-supervised learning. We discuss thepossibility that this result may be influenced by cognitive bias in LIDC-IDRIbecause malignancy estimates were recorded by the same radiologist panel asbiomarkers, as well as future work to incorporate pathology information over asubset of study participants.

Lung Nodule Classification Using Biomarkers, Volumetric Radiomics and 3D CNNs