NiBabel is a python package, under the NiPy project, that aims at unifying the process of opening different medical and neuroimaging file formats, including: ANALYZE,GIFTI, NIfTI1, MINC, MGH and ECAT as well as PAR/REC.
NiPy is an effort to make brain imaging research easier and more clear. This is implemented by providing a series of software that deal with file IO, analysis, and interfaces & pipelines.
BrainBrowser is a collection of open source, web-based 3D data visualization tools, mainly for neuroimaging studies. It is built using open technologies such as WebGL and HTML5. It allows exploration of cortical surface models (MNI and Wavefront OBJ, as well as FreeSurfer ASCII surface format) and volumetric MINC data.
Fiji is a distribution of ImageJ. The idea of the developers is to make the life of scientists easier by bundling ImageJ with nicely organised plugins and auto update function.
PySpace is a signal processing and classificiation environment for Python.
Modular software for processing of large data streams that has been specifically designed to enable distributed execution and empirical evaluation of signal processing chains.
Google has packaged their deeplearning machine learning tools and made it open source. The project is called tensorflow, and is available here. Some nice tutorials on the website, so that with a bit of patience, people can start to deep their toes into machine learning!