Home - Romain-Laine/MiLeSIM GitHub Wiki

Welcome to the MiLeSIM wiki! This wiki describes the content of the MiLeSIM package. Requirements: To run this package, it is necessary to have the machine learning and the neural network toolboxes with AlexNet, and the bio-formats MATLAB package as part of MATLAB 2017b.

The MiLeSIM package contains MATLAB m-files to perform:

1. automated extraction of particles from a fluorescence image (Otsu thresholding followed by active contour) (ExtractObjects_fromImage_BATCH.m)
2. automated extraction of the descriptors necessary for MiLeSIM classification (structural from masks, intensity based, image moments, bag of features and AlexNet, see paper for details). (ExtractDescriptors_BATCH.m)
3. automated classification of particles (ML_PredictWithModel_RF_BATCH.m)
4. automated structural analysis of spherical, filamentous and rod-like particles (PostAn_AnalyseStructures_BATCH.m)

NB: all the BATCH versions allow for batch analysis of multiple files/folders.