Reproduce “Parallel Multi-Dimensional LSTM, With Application to Fast Biomedical Volumetric Image Segmentation” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Marijn F. Stollenga, Wonmin Byeon, Marcus Liwicki, Juergen Schmidhuber
PyraMiD-LSTM segments brain tissue in three-dimensional MRI volumes by reorganizing multidirectional recurrent computation for greater parallelism while retaining context across the volume. Its design uses six directional LSTMs rather than the eight in a standard MD-LSTM. On the MRBrainS13/ISBI NEATBrain15 challenge, it ranked first overall, with CSF Dice overlap of 83.72%; this result is limited to the benchmark and setup.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.