You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Lesional volume is a powerful and commonly used biomarker in Brain MRI
analysis and interpretation.
Volume estimation is usually carried out through image segmentation, relying on Deep Convolutional Neural Networks (CNNs) trained on an annotated
database, comprising both images and their corresponding manual delineations.
To help clinicians focus on the more subtle regions, we propose to associate quantitative Predictive Intervals (PIs) for Volume estimations.
It is made out of a Research paper which was published around 23rd September.
What we tried to do is to take a step towards improvement and use up a New Pooling Technique called Stochastic Pooling(from yet another paper published around 2013).
All the Research Papers that are used are added in this Repo for Reference.