Magnetic resonance-based eye tracking using deep neural networks
DeepMReye is an open-source deep-learning framework for eye tracking in fMRI without camera.
It reconstructs viewing behavior from the MR-signal of the eyeballs. It works even in existing datasets and when the eyes are closed.
The Paper, Code, Data
& User Documentation are all open access!
Related, here are a few Tips & Tricks for improving eye-tracking quality in neuroimaging environments.
Functional imaging of the human medial temporal lobe.
A neuroscientist's guide to fMRI pulse sequence optimization
The medial temporal lobe (MTL) is difficult to image with fMRI due to magnetic field inhomogeneities and low signal-to-noise ratios.
Here, I compiled some information about fMRI pulse sequences and how they affect your data, along with a few tips on how to get a good signal in the MTL.
Find it here on Open Science Framework!
Behavioral encoding modeling
This code creates, fits & tests an encoding model of virtual head direction using simulated fMRI voxel time courses.
It can be easily adapted for other behavioral domains & imaging techniques (MEG, 2p-imaging...) to study the neural underpinnings of behavior across species.
