Development of machine learning-based regression models to predict inter-individual differences from MEG-based resting-state functional connectivity - Master thesis

Aug 2, 2026

Upload by Liam

Functional connectivity profiles provide a unique fingerprint for identifying individuals, enabling brain connectomics to study inter-individual variability during resting-state and task-evoked activity. This thesis investigates and predicts motor-task inter-individual differences using magnetoencephalography (MEG) resting-state functional connectivity. Regression-based models are developed and compared to improve identification on 51 Human Connectome Project subjects, and a residualisation approach increases across-subject variability. Results show accurate prediction of connectome differences for spontaneous and task-evoked activity, but no improvement in identification when incorporating muscle activity (EMG), indicating limits in predicting muscular activity from individual brain connectivity.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages118

Reading Time4 h 57 min

AccessFree to download

1 Introduction

2 Background

2.1 Brain anatomy and physiology

2.2 Magnetoencephalography

2.3 Human neuromuscular system

2.4 Mathematical tools

3 Materials and methods

3.1 Human Connectome Project

3.2 MEG pre-processing

3.3 EMG pre-processing

Master of Science in ICT for Internet and Multimedia Development of machine learning-based regression models to predict inter-individual diﬀerences from MEG-based resting-state functional connectivity Master Candidate Enrico Fongaro Supervisor Prof. Giulia Cisotto Student ID 2017403 Co-supervisors Prof. Viviana Betti Dr . Ottavia Maddaluno Academic Year 2022/2023 April, 3rd 2023 Abstract Recent research has demonstrated that functional connectivity proﬁles act as a unique ﬁngerprint that can accurately identify subjects from a large group. Asa result, brain connectomics has emerged as a rapidly growing research ﬁeld that focuses on identifying individuals based on inter-individual variability in brain connectivity during resting-state and task-evoked responses. The main objective of this study, which is part of the ERC HANDmade project (SH4, ERC-2017-STG) led by Prof. Betti Viviana, is to investigate and predict inter-individual diﬀerencesin motor tasks based on resting-state functional connectivity measured with magnetoencephalography (MEG) . Several regression-based models are discussed and compared to improve the identiﬁcation rate on 51 subjects from the Human Connectome Project. A residualisation approach is also presented to enhance model performance by increasing variability across subjects. Results indicate that interindividual diﬀerences in brain connectomes during spontaneous and task-evoked activity can be accurately predicted. However, it is observed that no improvement can be obtained in the identiﬁcation rate considering the muscle activity, compared to Vettoruzzo’s MSc thesis. Regardless of the model complexity and the number of the EMG features considered, the results show the inadequacy of predicting muscular activity from individual brain connectivity. Therefore, the investigation is addressed to understand the relationship between the brain activity and muscular activity, and non-linear approaches employing neural networks are considered the best choices for this type of motor tasks. The study’s ﬁndings can have implications in clinical research and rehabilitation ﬁelds for comprehending the neural mechanism underlying various neurological and muscular disorders and for developing personalized treatment approaches. Sommario Recenti ricerche hanno dimostrato che i proﬁli di connettività funzionale agiscono come una impronta digitale univoca in grado di identiﬁcare con precisionei soggetti di un gruppo. Di conseguenza, la connettomica, lo studio delle connessioni tra aree diverse del cervello umano, è emersa come un campo di ricercain rapida crescita che si concentra sull’identiﬁcazione della variabilità interindividuale nella connettività cerebrale, durante lo stato di riposo e le risposte evocate dal task. L’obiettivo principale di questo studio, che fa parte del progetto ERC HANDmade (SH4, ERC-2017-STG) coordinato dalla Prof.ssa Betti, è indagare e predire le diﬀerenze interindividuali nei task motori a partire dalla connettività funzionale spontanea, calcolata a partire da dati acquisiti tramite magnetoencefalograﬁa (MEG) . Diversi modelli basati sulla regressione vengono discussie confrontati per migliorare il tasso di identiﬁcazione su 51 soggetti del Human Connectome Project. Viene inoltre presentata una tecnica di residualizzazione per migliorare le prestazioni dei modelli aumentando la variabilità interindividuale nelle attività motorie. I risultati indicano che le diﬀerenze interindividuali nei connettomi cerebrali durante l’attività spontanea ed evocata dal task motorio possono essere predette con precisione. Tuttavia, si osserva che non si ottiene alcun miglioramento, rispetto ai risultati della tesi di laurea magistrale di Vettoruzzo, nel tasso di identiﬁcazione considerando l’attività muscolare. Indipendentemente dalla complessità del modello e dal numero delle features EMG considerate, i risultatimostrano l’inadeguatezza della previsione dell’attività muscolare a pa

FAQs

What is the main goal of this study?

To investigate and predict inter-individual differences in motor tasks using resting-state functional connectivity measured with MEG.

Which models are used to predict individual differences?

The work discusses and compares regression-based models, and it introduces a residualisation approach to enhance model performance by increasing variability across subjects.

How does the thesis assess the role of muscle activity (EMG)?

It evaluates identification and prediction including muscle activity features, but finds no improvement compared with a prior MSc thesis, showing the inadequacy of predicting muscular activity from individual brain connectivity.

Aug 2, 2026

Upload by Liam