FAST (v1.0.0)

The Feature-Assisted Segmenter/Tracker Have you struggled with automated measurement of the movement of many closely-packed individuals? The Durham lab's tracking system, FAST, may be the answer to your problems. Features A user-friendly GUI, designed to minimise the number of parameters that must be chosen and maximise immediate user feedback for those that do. Flexible design, accommodating a range of different workflows and dataset types. Rapid algorithms with low memory requirements, allowing thousands of objects to be tracked simultaneously on a standard laptop computer. Integrated overlay and plotting systems, allowing rapid validation and visualisation of tracking data. No need for a Matlab license - FAST is also available as a free stand-alone application. Detailed documentation available at https://mackdurham.group.shef.ac.uk/FAST_DokuWiki/dokuwiki Please note that FAST depends on Bio-Formats for full functionality. To use FAST from the code available here, please download the latest version of Bio-Formats for Matlab and add the bioformats_package.jar file to your homePanel/bfmatlab folder. Alternatively, you can download a pre-packaged version of FAST from the wiki. A publication detailing the implementation and functionality of FAST is currently under preparation. ------------------------------------------------------------------------------------------------------------------------------------------------------- FAST v1.0.0 changelog Changed theoretical basis of tracking and division detection modules to an information-theoretic approach. Broadly similar to previous approach, but more formal and powerful. Trackability score is now provided as the amount of information per object available for object assignment between frames. Now fully account for possible correlations between features by making use of the complete covariance matrix, as opposed to just the variances (the diagonal elements). Also fixed several bugs in the division detection module. Visual improvements to several modules. Completed and improved the code for generating and displaying kymographs in the plotting module. Added a new option in the plotting module, the 'feature space' option. Also allow users to plot all sampled points in the 'joint distribution' option of the plotting module, as well as adjust point colour according to local density (useful for large datasets). Improved readability of axis and colorbar titles, including automated addition of dimensional symbols from metadata. Added support for custom feature extraction, through the 'customFeats.m' function. Stabilised visualisation of the watershed algorithm in the segmentation module. Added documentation to the DoubleFAST functions