In the last few years, we have seen a very fast growth in the use of fMRI techniques to explore brain anatomy and functions in vivo, providing meaningful information about normal and abnormal conditions. In particular, the use of functional imaging techniques finds an interesting application in the study of heritability, since fMRI can deeply aid our understanding of individual differences in brain functions, and in cognitive and behavioral development. The heritability is de fined as the proportion of variation in a measurement caused by genetic influences, not including the influences due to environmental effects. The majority of heritability studies have been performed on twins, since their similar design can reveal whether specifi c neuroanatomical measures are predominantly influenced by genetics or environment, by comparing twin pairs with different degrees of genetic affinity. In this context, our work utilizes the twin method to study the heritability and environmental influences on resting-state connectivity patterns, aimed to investigate for a higher similarity of activation maps in monozygotic (MZ) twins with respect to dizygotic (DZ) twins, by using a multimodal analytical approach. The study considers data coming from 56 subjects which include 14 MZ (mean age: 16.5 ± 3.8) and 14 DZ pairs (mean age: 16.9 ± 4.9). Starting from their resting-state fMRI (rs-fMRI) data, functional connectivity has been computed by using two different approaches: Independent Component Analysis (ICA) based connectivity and Seed-based connectivity. In the first part of the analysis, ICA has been applied to the preprocessed rs-fMRI data in order to decompose them into 40 Independent Components (ICs). Among all these components, we kept only the ones relative to four networks of interest: default-mode network (DMN), visual, auditory and salience networks. We extracted the IC maps, related to each network, which best matched the corresponding resting-state template and used these in the comparison between activation patterns of MZ and DZ twins. The results have shown a signi cantly similarity in MZ compared to DZ in the visual network, but also the general trend supports the hypothesis that brain functional networks are at least partially genetically determined. The second part of our research was focused on the investigation of intra-pair similarity across MZ and DZ of functional connectivity patterns by exploring the whole brain volume with the Seed-based connectivity approach. We have also investigated the effective connectivity patterns on selected ROIs by means of Granger Causality Analysis (GCA). Through the extraction of Blood Oxygenation Level Dependent (BOLD) signal in correspondence of 84 Regions of Interest (ROIs) of an anatomical atlas, we built the brain network that was then used to perform the cross-correlation between pairs of ROIs for each subject. The fi nal aim was to assess intra-pair similarity and intra-pair difference. A signi cantly higher similarity in MZs was found in six pairs of ROIs involving also the occipital area. In addition, the similarity investigation was performed also on the local and global topological parameters of the brain network. Signi cant results of a lower intra-pair difference in MZ respect to DZ were found only in the Nodal Positive strength parameter in four specifi c ROIs. In the end, in order to provide information on the similarity of effective connectivity patterns in MZ with respect to DZ twins, we applied the GCA on the BOLD signal of six selected ROIs. However, the results obtained were considered associated only to environmental factors. In conclusion, although the current thesis was based on a preliminary twin dataset, we believe that the results obtained can provide interesting insights into genetic and environmental effects on brain networks. By enlarging the research to discordant twins, for example, it could be possible to identify preventive diagnostic protocols regarding the onset of pathological events.
Negli ultimi anni abbiamo assistito ad una rapida crescita nell'utilizzo di tecniche di imaging funzionale (fMRI), con lo scopo di esplorare l'anatomia e le funzioni cerebrali in vivo. L'uso della risonanza magnetica funzionale trova un'interessante applicazione nello studio dell'ereditabilità dei caratteri umani, sia fisiologici che patologici. Essa, infatti, è in grado di fornire maggiore chiarezza sulle differenze individuali e sullo sviluppo cognitivo e comportamentale a livello cerebrale. L'ereditabilità di un certo carattere di interesse è de finita come proporzione della varianza fenotipica dovuta ad influenze genetiche, al netto delle influenze dovute agli effetti ambientali. La maggior parte degli studi sull'ereditabilità sono stati eseguiti su coppie di gemelli, poichè il loro simile design genetico, e presumibilmente ambientale, ne può consentire una stima ottimale. A questo proposito, il nostro lavoro sfrutta il metodo gemellare per studiare ereditabilità e influenze ambientali sui patterns di connettività funzionale acquisiti in resting-state. Inoltre, si vogliono fornire prove statistiche di una più marcata somiglianza delle mappe di attivazione nei gemelli monozigoti (MZ) rispetto ai gemelli dizigoti (DZ), utilizzando un approccio analitico multimodale. Lo studio considera i dati provenienti da 56 soggetti che includono 14 coppie di MZ (età media: 16.5 ± 3.8) e 14 coppie di DZ (età media: 16.9 ± 4.9). A partire dai dati fMRI a riposo (rs-fMRI), la connettività funzionale è stata eseguita utilizzando due diversi approcci: connettività basata sull'Analisi delle Componenti Indipendenti (ICA) e connettività basata su seed/voxel. Nella prima parte dello studio, è stata applicata l'ICA ai dati rs-fMRI pre-elaborati al fine di decomporli in 40 Componenti Indipendenti (ICs). Di queste, abbiamo estratto solo quelle relative a quattro aree cerebrali di interesse: default-mode network (DMN), rete visiva, uditiva ed emotiva. L'estrazione delle mappe spaziali indipendenti, relative a ciascuna network, è stata effettuata in base alla maggiore somiglianza con un template di attivazione in resting-state, ed utilizzate successivamente per il confronto tra gemelli MZ e DZ. I risultati ottenuti hanno mostrato una somiglianza signi cativamente maggiore nei MZ solo nella rete visiva, ma la tendenza generale dei pattern di attivazione nei due gruppi di gemelli riflette lo stesso andamento. La seconda parte della nostra ricerca si è concentrata sull'indagine di similarità ed ereditabilità intra-coppia nei pattern di connettività funzionale ottenuti esplorando l'intero volume cerebrale con un approccio basato su seed/voxel. Inoltre, un'analisi di similarità è stata effettuata anche sui pattern di connettività effettiva estratti da sei ROIs di interesse tramite la Granger Causality Analysis (GCA). Attraverso l'estrazione del segnale BOLD (Blood Oxygenation Level Dependent) in corrispondenza di 84 Regioni di interesse (ROIs) di un atlante anatomico, abbiamo costruito la rete cerebrale che è stata poi utilizzata per eseguire la correlazione incrociata tra coppie di ROI per ciascun soggetto. L'obiettivo finale era valutare la somiglianza tra coppie in termini di cross-correlazione tra ROIs. I risultati di questa analisi hanno mostrato una somiglianza signi cativamente più alta nei MZ rispetto ai DZ in sei coppie di regioni che coinvolgono anche l'area occipitale. Inoltre, le indagini di similarità e ereditabilità sono state eseguite anche sui parametri topologici locali e globali estratti dalla rete cerebrale. In questo caso, risultati signi cativi coerenti con la nostra tesi sono stati trovati solo nel parametro noto come "forza nodale positiva" in quattro ROIs specifi che. In aggiunta, sono state stimate le interazioni causali tra sei regioni di interesse, scelte sulla base di risultati ottenuti in letteratura, tramite la GCA applicata ai segnali BOLD delle suddette sei aree. Tuttavia, i risultati ottenuti da questa analisi sono stati attribuiti a fattori casuali e non genetici. In conclusione, le evidenze estratte dalla presente ricerca, anche se basata su un dataset limitato, intendono fornire interessanti spunti sugli effetti genetici ed ambientali sulle reti cerebrali. Infatti, questo lavoro tende a porsi come tassello di una ricerca scientifi ca che, applicando ed affinando i metodi qui descritti su coppie di gemelli discordanti e non, possa identifi care protocolli diagnostici preventivi riguardo all'insorgenza di eventi patologici.
Brain functional connectivity analysis in couples of monozygotic and dizygotic twins
TORRE, TEODOLINDA
2017/2018
Abstract
In the last few years, we have seen a very fast growth in the use of fMRI techniques to explore brain anatomy and functions in vivo, providing meaningful information about normal and abnormal conditions. In particular, the use of functional imaging techniques finds an interesting application in the study of heritability, since fMRI can deeply aid our understanding of individual differences in brain functions, and in cognitive and behavioral development. The heritability is de fined as the proportion of variation in a measurement caused by genetic influences, not including the influences due to environmental effects. The majority of heritability studies have been performed on twins, since their similar design can reveal whether specifi c neuroanatomical measures are predominantly influenced by genetics or environment, by comparing twin pairs with different degrees of genetic affinity. In this context, our work utilizes the twin method to study the heritability and environmental influences on resting-state connectivity patterns, aimed to investigate for a higher similarity of activation maps in monozygotic (MZ) twins with respect to dizygotic (DZ) twins, by using a multimodal analytical approach. The study considers data coming from 56 subjects which include 14 MZ (mean age: 16.5 ± 3.8) and 14 DZ pairs (mean age: 16.9 ± 4.9). Starting from their resting-state fMRI (rs-fMRI) data, functional connectivity has been computed by using two different approaches: Independent Component Analysis (ICA) based connectivity and Seed-based connectivity. In the first part of the analysis, ICA has been applied to the preprocessed rs-fMRI data in order to decompose them into 40 Independent Components (ICs). Among all these components, we kept only the ones relative to four networks of interest: default-mode network (DMN), visual, auditory and salience networks. We extracted the IC maps, related to each network, which best matched the corresponding resting-state template and used these in the comparison between activation patterns of MZ and DZ twins. The results have shown a signi cantly similarity in MZ compared to DZ in the visual network, but also the general trend supports the hypothesis that brain functional networks are at least partially genetically determined. The second part of our research was focused on the investigation of intra-pair similarity across MZ and DZ of functional connectivity patterns by exploring the whole brain volume with the Seed-based connectivity approach. We have also investigated the effective connectivity patterns on selected ROIs by means of Granger Causality Analysis (GCA). Through the extraction of Blood Oxygenation Level Dependent (BOLD) signal in correspondence of 84 Regions of Interest (ROIs) of an anatomical atlas, we built the brain network that was then used to perform the cross-correlation between pairs of ROIs for each subject. The fi nal aim was to assess intra-pair similarity and intra-pair difference. A signi cantly higher similarity in MZs was found in six pairs of ROIs involving also the occipital area. In addition, the similarity investigation was performed also on the local and global topological parameters of the brain network. Signi cant results of a lower intra-pair difference in MZ respect to DZ were found only in the Nodal Positive strength parameter in four specifi c ROIs. In the end, in order to provide information on the similarity of effective connectivity patterns in MZ with respect to DZ twins, we applied the GCA on the BOLD signal of six selected ROIs. However, the results obtained were considered associated only to environmental factors. In conclusion, although the current thesis was based on a preliminary twin dataset, we believe that the results obtained can provide interesting insights into genetic and environmental effects on brain networks. By enlarging the research to discordant twins, for example, it could be possible to identify preventive diagnostic protocols regarding the onset of pathological events.| File | Dimensione | Formato | |
|---|---|---|---|
|
2018_07_Torre.pdf
non accessibile
Descrizione: Testo della tesi
Dimensione
4.57 MB
Formato
Adobe PDF
|
4.57 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/141637