Scale Dependent Convolution layer

Dealing with scale with parametrized dilated convolution

Motivation

Image analysis pipelines based on the segmentation, extraction of features of interest and aggregation of this information for subsequent prediction, allow for apriori and aposteriori analysis in addition to being more “trustworthy”. Indeed as the model is build around objects selected by experts, predictions can be traced back to these elements, thus increasing the trust and interpretability.

Downloading histopathology data from the TCGA

How to download TCGA samples with a script and a manifest.

The Cancer Genome Atlas, a landmark cancer genomics program, molecularly characterized over 20,000 primary cancer and matched normal samples spanning 33 cancer types. This joint effort between the National Cancer Institute and the National Human Genome Research Institute began in 2006, bringing together researchers from diverse disciplines and multiple institutions.

ISBI 2019

Publication in ISBI 2019 Predicting Residual Cancer Burden in a triple negative breast cancer cohort.

At the ISBI 2019 held in Venice, Italy, I presented our conference paper Predicting Residual Cancer Burden in a triple negative breast cancer cohort during the poster session. The poster can be found here.

Deep Learning For Image Analysis

I was a teaching assistant for the segmentation practical sessions for the course Deep Learning For Image Analysis given in the ATHENS Programme by Etienne Decencière, Santiago Velasco-Forero and Thomas Walter. I animated a practical session on tissue segmentation, the notebook and presentation can be found in github and the data can be found here.

Deep Distance Regression

Publication in IEEE Transactions on Medical Imaging of Segmentation of Nuclei in Histopathology Images by deep regression of the distance map.

The paper ‘Segmentation of Nuclei in Histopathology Images by deep regression of the distance map’ by Peter Naylor, Thomas Walter, Fabien Reyal and Marick Laé has been published in IEEE transactions on medical imaging, 2018.

Pagination