GitHub Repositories and Python packages

Here is a non exhausted list of informatics projects I have contributed to:

• ISRIN, Code to reproduce the work in Implicit Neural Representation for Ice Sheet Surface Elevation Reconstruction by Peter Naylor, Andreas Stokholm, Natalia Havelund Andersen, Nikolaos Dionelis, Quentin Paletta, Sebastian Bjerregaard Simonsen.
• SmRMR, Code to reproduce the results of the paper Sparse minimum Redundancy Maximum Relevance for feature selection by Peter Naylor, Benjamin Poignard, Hector Climente-Gonzales and Makoto Yamada. Implements a Sparse Minimum Redundancy Maximum Relevance that can be used in the style of scikit-learn. Writen in Jax.
• INR4torch, python package and in development. Basically a toolkit for using Implicit Neural Representation and Physics Informed Neural Networks with Python.
• NN-4-CD, code to reproduce the work Implicit neural representation for change detection by P. Naylor, D. Di Carlo, A. Traviglia, M. Yamada and M. Fiorucci.
• ScaleDependentCNN, Scale Dependent Convolutional layer for feature extraction of nuclei in histopathology data. The code for the pipeline is written in Nextflow and the processes use python.
• Galore, Graphical models applied to gene-expression data. Main author is Hector Climente-Gonzalez (on going).
• Knockoff-MMD-HSIC, python package and reproducible code linked to a submitted paper. Code co-developed with Hector Climente-Gonzalez (01/11/2021)
• NucSeg: Nextflow pipeline to train a nuclei segmentation network with all available nuclei segmentation datasets (at the time) with model such as ResNet on steroids and so on. (01/09/2021)
• The dart Application which is a flask/dash website for custom dart games hosted on our family Raspberry Pi. Project developed with Bruno Naylor (26/06/2021).
• PhD_template: Template used for my dissertation (17/12/2019).
• Cellular heatmap generation and classification: A project based on the segmentation of nuclei in histology images, from the segmentation we extract a certain number of features, perform an unsupervised dimension reduction on the nuclei sets, transform this extraction to the tissue in order to generate cellular heatmaps, and finally a supervised model for the classification of these heatmaps (12/11/2019) (on going).
• AutomaticWSI: The prediction of residual cancer burden from histological TNBC biopsy sections prior to treatment, the code contains a full pipeline to analyze Whole Slide Images automatically (8/12/2019) (on going).
• BestNucleiModel: A nested-cross validation distributed with Nextflow to assess the best nuclei segmentation model (10/12/2018).
• segmentation_net package: Python code based on Tensorflow for segmentation tasks (not available yet).
• dynamic_watershed package: Python package implementing the algorithm for splitting nuclei described in in ‘Nuclei segmentation in histopathology images using deep neural networks’ (27/08/2018).
• useful_wsi package: Python code implementing a set of useful tools to deal with Whole Slide Image (WSI) formats (14/07/2018).
• Segmentation of Nuclei in Histopathology Images by deep regression of the distance map: This Github repository contains the necessary code to reproduce the work contained in the submitted paper (cf. link name) (24/04/2018).