Internship: Computational methods in Neuroscience

Student Project
Amsterdam - Remote
Inter-subject correlation and hyperalignment of functional magnetic resonance imaging data. Structural and statistical analysis of functional ultrasound imaging data. Machine learning and deep learning.
4 - 12 months
Daily support
Lorenzo de Angelis


At the Social Brain Lab we strive for applying the most advanced analysis strategies to our data. If you are an excellent student with a passion for computation and mathematical modeling you might be interested in taking part in one of our projects. The data collected in our lab typically involves the understanding of emotions and their neural correlates. The type of data available for these projects ranges from fMRI to electrophysiological recordings as well as novel techniques such as functional ultrasound imaging (fUSI).

What we look for

  • Excellent skills in analytical disciplines (math, physics, statistics, ….) and affinity with programming languages (python, R, …);
  • Fast-paced learner who is not afraid to develop analytical methods that are not implemented in standard packages;
  • Neuroscience enthusiast, excited about working with experimental neural recordings.

What we offer

  • Opportunity for an experience as part of a multidisciplinary research team consisting of a wide variety of scientists with diverse expertise;
  • A balanced supervision with room to develop your own ideas;
  • A friendly collaborative environment.

Potential topics for projects

Inter-subject correlation
Emotions are not something that we can easily turn on and off like a light switch. This intrinsic naturalistic character of emotional stimuli makes it difficult to model neural data that was collected while exposing subjects to emotional content. A possible solution is to use one person’s response as a model for others, by correlating the brain activity across individuals that witnessed the same event.

When comparing the brain response of different subjects (humans or animals) we have to deal with the fact that every brain is different, in shape, size and functionalities. A successful approach that has been pursued for many years in fMRI and is still the gold standard for many applications, is to match the morphology of white/gray matter to match a given brain template. This of course exhibits limitations, as some features of the brain might develop differently in different subjects as they age. A solution that has been advocated to circumvent the limitations of this anatomical registration is the use of hyperalignment, which uses the functional activity rather than the anatomical morphology to align the brains of different individuals.

Functional Ultrasound Imaging
Functional ultrasound imaging (fUSI) is a novel brain imaging technique that has the potential for becoming a highly portable and non-invasive way to infer brain activity. At the Social Brain Lab, we set up an ultrasound imaging facility for imaging the brain of head-fixed rodents. This technique offers a hemodynamic readout reminiscent of fMRI, making it an important bridge for connecting human and rodent research. For this project there are opportunities both for processing of the imaging data and/or for involvement in the technical experimental aspects of fUSI.

What next

If you think you are a good match and you are interested in one of these topics or want to send a spontaneous application, please contact our analytics expert Lorenzo De Angelis at l.deangelis [at]