Tipo di tesi
Tesi di laurea magistrale
Titolo
Detection and statistical testing of cell assemblies from multivariate continuous time series
Parole chiave
- cell assembly detection
- continuous time series
- inhibition
- overlapping assemblies
Data inizio appello
22/06/2026
Consultabilità
Non consultabile
Data di rilascio
22/06/2029
Riassunto (Inglese)
Cell assemblies are groups of neurons, whose coordinated activity is linked to retrieval of a specific concept or to the execution of actions. The aim of this thesis is to develop and test a new algorithm for the unsupervised detection of cell assemblies from both continuous and discrete experimental time series. Unlike previous methods, it can identify a broad range of assembly patterns across multiple timescales, without requiring the a priori setting of arbitrary parameters or relying on slow bootstrap significance testing. Through the work done in this thesis, its key strengths now include the ability to detect overlapping assemblies and to identify inhibitory-excitatory relationships between assembly units. The proposed algorithm was successfully validated on both simulated and experimental datasets, proving useful for the analysis of neural recordings across multiple regions, especially in the presence of inhibitory relationships or shared neurons.
