Vés enrere

Thesis Defence: Irene Malvestio

Thesis Defence: Irene Malvestio



Irene Malvestio

Detection of directional interactions between neurons from spike trains

Supervisors: Dr. Ralph G. Andrzejak and Dr. Roberto Livi (Università di Firenze)



Thesis Brief Description:

An important problem in neuroscience is the assessment of the connectivity between neurons from their spike trains. In my thesis I study an approach for the detection of directional couplings between dynamics based on recorded point processes. This method, called nonlinear interdependence measure L, was developed to extend the classical linear approaches, in order to better deal with nonlinear couplings and directional detection of
interactions. In my work I first test the method with simulated data and compare its performances against the one of linear measures. Finally, I apply all measures to neuronal spiking data from an intracranial whole-night recording of a patient with epilepsy. The network reconstructions that I get proved to be coherent during the same sleep stage in different moment of the recording session.

Experience as a PhD Student:

Doing a PhD is very hard but is also a great learning experience. I was part of a Marie Curie Innovative Training Network, so I got amazing opportunities to collaborate with other scientists from different countries. These helped me broadening my research interests, and better understand how science works. Conferences gave me the opportunity to meet people and enjoy travelling. My colleagues have been great friends during the whole experience! My supervisors helped me in learning to deal with the difficulties of research, to be confident in expressing my opinion, and finally, they taught me how to be more independent. Last but not least, living in a vibrant city like Barcelona gives me the opportunity to enhance my creativity with great dancing and music lessons.