Elsevier's profit margin compared to Apple, Google, and Microsoft

Apple: 28%
Google: 25%
Microsoft: 34%

Elsevier: 37% with a revenue of $3.9 billion.

Elsevier's payment to academic authors and reviewers: $0

In any case, I have officially awarded myself with best poster🏆

Thanks for your kind response. Do you think these brain-wide activation could be related to the action-mode network? The AMN seems to be a global state in order to plan goal-direction behavior.

We'd love to hear your thoughts!

Glad you like it! Note that this work is from my PhD work which I did at neuralinterfacinglab.github.io . I am currently a postdoc a BrainGate working on new projects on movement decoding using microelectrode arrays in the motor cortex!

And finally, science is a team-effort. We are incredible thankful for all collaborators providing their input and expertise, the invaluable staff at Kempenhaeghe and our participants for providing their time and effort.

We are excited about this work and hope you enjoy reading it. As we always do in our lab, we share code and data publicly. Please find the code here: github.com/mottenhoff/d.... We'll make the data available at publication.

Moreover, our results suggest that goal-directed movement in the brain is represented in relationship to that goal. Current brain-computer interfaces do already use this goal-directed reference frame to calibrate decoders. However, it does require knowledge of the location of the goal.

Our results provide a comprehensive overview of the decodable movement-related information in brain-wide neural activity, and strengthen our previous results that the whole brain may be involved in generating movement.

In this work, the decoder reached the highest performance using low frequency activity. The phase information is important, as the performance was substantially lower when we used the power. This neural signal shares similarities with the local motor potential described in M1.

About the decoding of the directional kinematics. It turns out that if we change the reference frame from the sensor to a goal-directed reference frame, the decoder is able the position w.r.t the target above chance!

If we look at the electrode contact locations that were significantly correlated with hand movement speed, we observe that these contacts are located throughout the brain, including cortical and subcortical structures.

We found that non-directional hand movement speed can be decoded using low, mid and high frequency information. Particularly, the low frequency activity reached substantial decoding performance (up to 0.76 + 0.03). Directional hand kinematics seemed not decodable.

We extracted low frequency activity (< 5 HZ), mid frequency power (8 - 30 Hz) and high-frequency power (55 - 200 Hz), windowed the data (300ms length, 50ms shift), and used preferential subspace identification (PSID) to decode 12 kinematics.

To investigate, we recorded 3D hand movement trajectory while our participants played a custom game from which we retrieved position, velocity, speed and acceleration. At the same time we recorded neural activity from cortical and sub-cortical brain areas.

In previous work, we described how movement can be decoded from brain-wide neural activity, across tasks and across participants. However, this was only decoding between move and rest, so naturally we wanted to know more about the neural content of these brain-wide dynamics.