In complex networks, information transfer is ensured by an efficient topological architecture. A methodology to causally validate this structure-function relationship is represented by Transcranial Magnetic Stimulation (TMS), a form of noninvasive brain stimulation that can act on the neural mechanisms underpinning cognition. The aim of this project was to study the topographical properties of the individual connectome, determine its response to external perturbations and use this information to tailor the selection of stimulation targets in the brain. To achieve these aims, we conducted a series of studies employing graph theory and network theory analyses to determine interindividual differences in the brain information flow and cognitive efficiency. In Study 1, we investigated how interindividual differences in brain topology account for differences in executive functions (EF) and how genetic factors might shape such a relationship. The neuropsychological and resting state functional magnetic resonance imaging data of 453 twins from the Colorado Longitudinal Twins’ Study and 463 twins from the Human Connectome Project (HCP) were analyzed. Through an approach of step-wise in silico network lesioning of the individual functional connectome, we showed that interindividual differences in EF are associated with different dependencies on neural networks at rest. In Study 2, we further addressed the heritability of brain resilience to in silico network lesioning, following the removal of either brain nodes or connections. Evidence of moderate heritability was found for inter-networks, interhemispheric links and for additional topological indices. In Study 3, we moved to investigate if personalized neurostimulation results in enhanced protocol’s reliability and efficacy. Twenty-four subjects underwent single pulse TMS over two nodes belonging to the Dorsal Attention (DAN) and Default Mode (DMN) Networks, respectively. Across visits, the stimulated target for both networks was chosen either based on group-derived networks’ maps or personalized based on individual anatomy and functional profile. All stimulation visits were conducted twice, one month apart, during concomitant electroencephalography recording. Preliminary results suggest higher reliability of the results following individualized protocols, especially for the DAN stimulation. Finally, in Study 4, we tested network control theory approaches for the prediction of the optimal stimulation target(s) in TMS interventions. Simulations were run on 400 HCP participants, with test-retest data available for 45 of them. Ideal stimulation nodes were defined as the ones able to guide the system from an initial state to a desired target state. We modeled the efficacy of stimulation applied to traditional stimulation sites compared to input nodes derived from network control theory (NCT) predictions. Results suggest that the amount of network engagement following stimulation of NCT-derived cerebral sites is significantly higher compared to traditionally employed neuromodulation sites, suggesting NCT as a useful tool in guiding brain stimulation. This project aims at studying interindividual differences in brain topology as a proxy to understand inter-subject variance in high order cognitive functioning and the impact of genetic and environmental influences in shaping this relationship. Furthermore, it represents a first attempt in addressing the importance of personalization of care in consideration of the benefit/cost ratio. Overall, this project highlights the strengths of applying network sciences to the understanding of the continuum between brain topology and cognitive functioning, of the modeling of information transfer in the brain, and in the neural pathways underpinning brain stimulation.
In network complessi, il passaggio dell’informazione si basa su un’efficiente organizzazione topologica. La stimolazione transcranica magnetica (TMS) è una forma di stimolazione cerebrale non-invasiva che ci permette di testare la causalità tra meccanismi neurali e funzionamento cognitivo. Lo scopo di questo progetto è stato quello di studiare le proprietà topografiche del connettoma individuale, determinare i suoi pattern di risposta a perturbazioni esterne ed utilizzare questa informazione per l’individualizzazione dei target di stimolazione cerebrale. A tale scopo, abbiamo condotto una serie di studi basati su nozioni derivanti dalla teoria dei grafi e analisi dei network per studiare le differenze interindividuali e comprendere come queste influenzino la performance cognitiva. Nello Studio 1, abbiamo studiato il rapporto tra topografia cerebrale, funzioni esecutive (EF) e predisposizioni genetiche. Abbiamo quindi analizzato i dati neuropsicologici e di risonanza magnetica funzionale di 453 gemelli provenienti dal Colorado Longitudinal Twins’ Study e 463 gemelli provenienti dall’Human Connectome Project (HCP). Per mezzo di simulazioni di lesioni al connettoma funzionale individuale, abbiamo dimostrato un’associazione tra differenze interindividuali alle EF e differenti pattern di predominanza nei network cerebrali a riposo. Nello Studio 2, abbiamo testato l’ereditabilità della resilienza cerebrale a seguito di simulazioni di tali lesioni, osservando così una moderata ereditabilità per connessioni inter-network ed interemisferiche, nonché per alcune misure topologiche. Nello Studio 3, abbiamo esaminato se approcci di neurostimolazione personalizzati risultassero in una maggiore efficacia e replicabilità dei protocolli. Ventiquattro partecipanti sono stati sottoposti ad un protocollo TMS a singolo impulso su due nodi cerebrali appartenenti ai Network del Dorsal Attention (DAN) e del Default Mode (DMN) rispettivamente. Per ogni visita, il target di stimolazione poteva essere scelto sulla base di mappe dei network derivanti da statistiche di gruppo, o poteva essere invece personalizzato. La stimolazione veniva ripetuta due volte, a distanza di un mese, durante la co-registrazione del segnale elettroencefalografico. I nostri risultati dimostrano una maggior replicabilità dei risultati a seguito della personalizzazione dei protocolli di stimolazione, specialmente per il DAN. Infine, nello Studio 4, abbiamo testato l’utilizzo di approcci basati sulla teoria della controllabilità dei network (NCT) per predire i siti di stimolazione ottimali negli interventi TMS. Abbiamo testato diverse simulazioni su 400 partecipanti dell’HCP, con misure ripetute per 45 di essi. I nodi di stimolazione ideali venivano definiti sulla base della loro capacità di guidare il sistema da uno stato di attivazione iniziale, verso un determinato stato finale. I nostri risultati suggeriscono che il reclutamento dei network cerebrali a seguito di stimolazioni guidate dall’approccio NCT era significativamente maggiore rispetto che a seguito della stimolazione di nodi tradizionali, suggerendo quindi l’uso della NCT come un approccio utile nel guidare la scelta dei siti di stimolazione. Questo progetto ha lo scopo di studiare le differenze interindividuali nella topologia cerebrale come veicolo per comprendere la variabilità interindividuale in funzioni cognitive e l’impatto che influenze genetiche e ambientali possono avere nel modellare questa interazione. Inoltre, rappresenta un primo tentativo verso la personalizzazione degli interventi di cura, considerato il rapporto tra costi e benefici. Nel complesso, questo progetto sottolinea l’importanza di applicare lo studio dei network per comprendere il continuum tra topologia cerebrale e funzionamento cognitivo, del creare modelli rappresentativi del passaggio dell’informazione nel cervello, così come dei meccanismi neurali sottostanti la stimolazione cerebrale.
Studio della topologia cerebrale individuale per la personalizzazione degli interventi di stimolazione transcranica magnetica: un approccio alla modulazione dei network / Menardi, Arianna. - (2022 Feb 14).
Studio della topologia cerebrale individuale per la personalizzazione degli interventi di stimolazione transcranica magnetica: un approccio alla modulazione dei network
MENARDI, ARIANNA
2022
Abstract
In complex networks, information transfer is ensured by an efficient topological architecture. A methodology to causally validate this structure-function relationship is represented by Transcranial Magnetic Stimulation (TMS), a form of noninvasive brain stimulation that can act on the neural mechanisms underpinning cognition. The aim of this project was to study the topographical properties of the individual connectome, determine its response to external perturbations and use this information to tailor the selection of stimulation targets in the brain. To achieve these aims, we conducted a series of studies employing graph theory and network theory analyses to determine interindividual differences in the brain information flow and cognitive efficiency. In Study 1, we investigated how interindividual differences in brain topology account for differences in executive functions (EF) and how genetic factors might shape such a relationship. The neuropsychological and resting state functional magnetic resonance imaging data of 453 twins from the Colorado Longitudinal Twins’ Study and 463 twins from the Human Connectome Project (HCP) were analyzed. Through an approach of step-wise in silico network lesioning of the individual functional connectome, we showed that interindividual differences in EF are associated with different dependencies on neural networks at rest. In Study 2, we further addressed the heritability of brain resilience to in silico network lesioning, following the removal of either brain nodes or connections. Evidence of moderate heritability was found for inter-networks, interhemispheric links and for additional topological indices. In Study 3, we moved to investigate if personalized neurostimulation results in enhanced protocol’s reliability and efficacy. Twenty-four subjects underwent single pulse TMS over two nodes belonging to the Dorsal Attention (DAN) and Default Mode (DMN) Networks, respectively. Across visits, the stimulated target for both networks was chosen either based on group-derived networks’ maps or personalized based on individual anatomy and functional profile. All stimulation visits were conducted twice, one month apart, during concomitant electroencephalography recording. Preliminary results suggest higher reliability of the results following individualized protocols, especially for the DAN stimulation. Finally, in Study 4, we tested network control theory approaches for the prediction of the optimal stimulation target(s) in TMS interventions. Simulations were run on 400 HCP participants, with test-retest data available for 45 of them. Ideal stimulation nodes were defined as the ones able to guide the system from an initial state to a desired target state. We modeled the efficacy of stimulation applied to traditional stimulation sites compared to input nodes derived from network control theory (NCT) predictions. Results suggest that the amount of network engagement following stimulation of NCT-derived cerebral sites is significantly higher compared to traditionally employed neuromodulation sites, suggesting NCT as a useful tool in guiding brain stimulation. This project aims at studying interindividual differences in brain topology as a proxy to understand inter-subject variance in high order cognitive functioning and the impact of genetic and environmental influences in shaping this relationship. Furthermore, it represents a first attempt in addressing the importance of personalization of care in consideration of the benefit/cost ratio. Overall, this project highlights the strengths of applying network sciences to the understanding of the continuum between brain topology and cognitive functioning, of the modeling of information transfer in the brain, and in the neural pathways underpinning brain stimulation.File | Dimensione | Formato | |
---|---|---|---|
tesi_definitiva_Arianna_Menardi.pdf
Open Access dal 15/02/2023
Descrizione: tesi_definitiva_Arianna_Menardi
Tipologia:
Tesi di dottorato
Licenza:
Altro
Dimensione
10.66 MB
Formato
Adobe PDF
|
10.66 MB | Adobe PDF | Visualizza/Apri |
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.