University of Jyväskylä

Dissertation: 24.11.2017 Data-driven Analysis for fMRI during Naturalistic Music Listening (Tsatsishvili)

Start date: Nov 24, 2017 12:00 PM

End date: Nov 24, 2017 03:00 PM

Location: Mattilanniemi, Agora, Beeta

Valeri Tsatsishvili picture: Manana Koberidze
FM Valeri Tsatsishvili defends his doctoral dissertation in Mathematical Information Technology "Data-driven Analysis for fMRI during Naturalistic Music Listening". Opponent Professor Karen Egiazarian (Tampere University of Technology) and custos Professor Tapani Ristaniemi (University of Jyväskylä). The doctoral dissertation is held in English.

Understanding how human brain achieves perception and cognition of our surrounding world has been active branch of cognitive neuroscience since the inception of non-invasive brain imaging techniques such as fMRI (functional magnetic resonance imaging). Traditionally, in controlled neuroimaging experiments, participants have been exposed to simplified sensory stimuli or performed trivial cognitive tasks in brain imaging devices. However, even very common everyday activities, such as listening of music, involves highly complex cognitive processing of multi-channel information. For example, while listening to music, our brain analyzes and integrates various aspects of music information, such as its genre, rhythm, melody, structure, and timbre. Neural processing of these musical features has been studied in isolation, by simplified and/or artificially generated stimuli. However, recently it has been questioned whether studying neural processing of isolated musical features in controlled experiments reveals the complete picture of how such processing and integration happens in real life.

Consequently, several studies have conducted naturalistic neuroimaging experiments, where participants continuously listened to complete pieces of music in the fMRI scanner. Analysis of brain responses elicited in such naturalistic experiments poses new methodological challenges associated to the complexity of the stimulus, lack of control over experimental variables, and increased noise levels in the signals.

The present study addressed some of these methodological challenges by applying data-driven analysis methods to real fMRI data obtained from the naturalistic music listening experiment. One of the challenges is to identify and separate brain responses related to different aspects of stimulus, or artifacts originating from other sources including physiological processes and the imaging equipment. This thesis proposes few approaches to improve the source separation process. Another part of this work explores application of nonlinear methods in generation of the features that represent timbral, tonal, and rhythmic aspects of music stimulus.

More information

Valeri Tsatsishvili, 0443518343,,

Communications Manager Liisa Harjula, 040 8054403,

Valeri Tsatsishvili completed high school in 2000 in Tbilisi, Georgia. He earned his bachelor's degree in Physics from Iv. Javakhishvili Tbilisi State University, in 2004 and Master's degree from the University of Jyväskylä in 2011.

Jyväskylä Studies in Computing 268, 51 s., Jyväskylä 2017, ISSN: 1456-5390, 268, ISBN: 978-951-39-7239-4 (nid.), 978-951-39-7240-0 (PDF).


Interest towards higher ecological validity in functional magnetic resonance imaging (fMRI) experiments has been steadily growing since the turn of millennium. The trend is reflected in increasing amount of naturalistic experiments, where participants are exposed to the real-world complex stimulus and/or cognitive tasks such as watching movie, playing video games, or listening to music. Multifaceted stimuli forming parallel streams of input information, combined with reduced control over experimental variables introduces number of methodological challenges associated with isolating brain responses to individual events.

This exploratory work demonstrated some of those methodological challenges by applying widely used data-driven methods to real fMRI data elicited from continuous music listening experiment. Under the general goal of finding functional networks of brain regions involved in music processing, this work contributed to improvement of the methodology from two perspectives. One is to produce a set of representative features for stimulus audio that can capture different aspects of music, such as timbre and tonality. Another is to improve reliability and quality of separation of the observed brain activations into independent spatial patterns. Improved separation in turn enables better differentiation of stimulus-related activations from the ones originating from unrelated physiological, cognitive, or technical processes.

More specifically, part of the research explored an application of a nonlinear method for generating perceptually relevant stimulus features representing high-level concepts in music. Another part addressed dimensionality reduction and model order estimation problem before subjecting fMRI data to source separation and offered few methodological developments in this regard.

More information

FM Valeri Tsatsishvili
Filed under: