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

Prognostic tumor growth modeling via volumetric medical imaging observationscan potentially lead to better outcomes of tumor treatment and surgicalplanning. Recent advances of convolutional networks have demonstrated higheraccuracy than traditional mathematical models in predicting future tumorvolumes. This indicates that deep learning-based techniques may have greatpotentials on addressing such problem. However, current 2D patch-based modelingapproaches cannot make full use of the spatio-temporal imaging context of thetumor's longitudinal 4D (3D + time) data. Moreover, they are incapable topredict clinically-relevant tumor properties, other than volumes. In thispaper, we exploit to formulate the tumor growth process through convolutionalLong Short-Term Memory (ConvLSTM) that extract tumor's static imagingappearances and capture its temporal dynamic changes within a single network.We extend ConvLSTM into the spatio-temporal domain (ST-ConvLSTM) by jointlylearning the inter-slice 3D contexts and the longitudinal or temporal dynamicsfrom multiple patient studies. Our approach can incorporate other non-imagingpatient information in an end-to-end trainable manner. Experiments areconducted on the largest 4D longitudinal tumor dataset of 33 patients to date.Results validate that the ST-ConvLSTM produces a Dice score of 83.2%+-5.1% anda RVD of 11.2%+-10.8%, both significantly outperforming (p<0.05) other comparedmethods of linear model, ConvLSTM, and generative adversarial network (GAN)under the metric of predicting future tumor volumes. Additionally, our newmethod enables the prediction of both cell density and CT intensity numbers.Last, we demonstrate the generalizability of ST-ConvLSTM by employing it in 4Dmedical image segmentation task, which achieves an averaged Dice score of86.3+-1.2% for left-ventricle segmentation in 4D ultrasound with 3 seconds perpatient.

Spatio-Temporal Convolutional LSTMs for Tumor Growth Prediction by Learning 4D Longitudinal Patient Data