BioSMART will be the place where researchers from different Bio-engineering fields can meet and exchange ideas to build outstanding and innovative future projects. All accepted papers are expected to be included in IEEE Xplore and indexed by EI.

Delegates can register to attend BioSMART either On-site in Paris and there will be also an Online version.

Conference Objectives

• Bio-Signal and Image processing
• Bio-engineering systems
• Smart Technologies and AI applications

George Marinos from the University of Piraeus Research Center present the iHelp project’s paper which will be published in the first track (~Bio-signal and image processing) because is closely related to the biomedical image processing field.

Paper Description

This paper demonstrates a technique that can efficiently handle and manipulate CT Scan images annotated with survival information for the purpose of risk group identification in a patient’s cohort (or in other words in a dataset). Given the fact that we have a set of biomedical CT Scan images with survival annotations, we can apply this technique to stratify the risk of survival and finally obtain clusters of patients that have similar CT Scans and also have similar survival distribution (in the same cluster) and have different CT Scans and their survival probability distribution differentiates (when they are in different clusters). This can be achieved by utilizing a Deep Convolutional Neural Network for Survival risk estimation in the way that is explained in the paper. Experiments in 3 different publically available datasets, as presented in the paper, have proven that this technique can be successfully used for survival risk stratification.