Leonardo Bonetti
Center for Music in the Brain
Aarhus University
Centre for Eudaimonia and Human Flourishing
Linacre College
University of Oxford
Combining cutting-edge neuroscience and mathematically controlled music paradigms to uncover the brain mechanisms of human memory
My research aims to understand the brain mechanisms underlying memory for temporal sequences using a combination of neuroimaging tools (e.g. magnetoencephalography and magnetic resonance imaging), musical memory paradigms and advanced analytical techniques.
I began my journey in music, earning the traditional 10-year diploma (now equivalent to a master’s degree) in Classical Guitar from the Conservatory of Bologna (2005-2013). During this time, while performing and memorising complex musical pieces for concerts, I became deeply intrigued by the psychology of memory. This curiosity led me to pursue both a bachelor's (2011-2014) and master’s in psychology (2014-2016) at the University of Bologna.
My interest soon expanded beyond the behavioural aspects of memory to understanding the underlying neural mechanisms. This prompted me to begin a PhD in cognitive neuroscience at the Center for Music in the Brain, Aarhus University (2017-2020), where I investigated how the brain encodes musical memory. Throughout my PhD, I developed a strong interest in data analysis and computational methods, which I deepened through research stays (e.g. at MIT in 2019) and during my subsequent Junior Research Fellowship at the University of Oxford (2021-2024).
This multidisciplinary path has allowed me to merge neuroscience with advanced analytical methods and musical paradigms, frequently drawing attention from both the media and the public, and earning me prestigious awards and grants such as the Lundbeck Foundation Talent Prize 2022. My research trajectory has also led to my appointment as Associate Professor at the Center for Music in the Brain, Aarhus University, and as a Research Fellow at the Centre for Eudaimonia and Human Flourishing, University of Oxford. At these institutions, I apply my expertise in music, psychology, neuroscience, and data science to lead a research program focused on human memory, predictive brain processes, and aging.
Throughout my career, I have primarily investigated the brain mechanisms underlying auditory memory and perception, aging, and cognitive functions using magnetoencephalography (MEG) and magnetic resonance imaging (MRI).
Brain dynamics and long-term recognition of auditory sequences
I uncovered hierarchical brain mechanisms during the encoding, recognition and prediction of auditory sequences, demonstrating interactions between the auditory cortices, hippocampus, and cingulate gyrus. I identified differential brain processing speeds for recognising previously memorised versus novel sequences, with slower processing linked to familiar sequences and faster processing to novel ones. My research showed that tonal and atonal musical sequences engage distinct neural networks, with tonal sequences involving the hippocampal and cingulate areas, and atonal sequences activating the auditory processing network. Additionally, I revealed that individuals with higher working memory abilities recruit a broader network, including visual processing areas, for successful auditory memory recognition.
Age-related changes in neurophysiology and predictive coding
I discovered that while older adults could recognise memorised auditory sequences as effectively as younger adults, their brain's functional organisation underwent significant reshaping. Older adults exhibited increased early activity in sensory regions, such as the left auditory cortex, and decreased activity in the medial temporal lobe, pointing to compensatory mechanisms. However, when processing varied sequences, older adults showed a reduced prediction error in the medial temporal lobe and lacked compensatory mechanisms, leading to poorer performance. These findings suggest that the neurophysiology of memory and sequence prediction provides critical insights into healthy aging beyond behavioural performance alone.
Automatic brain predictive processes
I discovered that automatic prediction error is modulated by a complex interplay between genetic (e.g., COMT and BDNF genes), environmental (e.g., musical training), and psychological (e.g., depression, working memory) factors. This research shed new light on prediction error, identifying MMN as a potential biomarker for disorders such as depression.
Cognitive abilities, brain networks and musical training
Computing graph theory measures on functional (from MEG resting state) and structural (from DTI) connectivity networks, I discovered that highly versus average intelligent individuals (high versus average Gf) presented a higher level of integration of information across the whole-brain and a more efficient segregation of local brain subnetworks.
I also revealed that musical training is associated with functional and structural changes in the brain, as indexed by modulated neurophysiological signals and connectivity. Moreover, I showed that musical practice is related to cognitive abilities, with higher training associated with better cognitive performance beyond the musical domain.
Elite football, cognitive abilities and personality
Football is arguably the most widely followed sport worldwide, and many dream of becoming soccer players. However, only a few manage to achieve this dream, which has cast a significant spotlight on elite football players who possess exceptional skills to rise above the rest. In our research, we investigated the psychological profile of elite football players, revealing that success on the field goes beyond physical ability. By analysing a sample of 328 participants, including 204 elite football players from the top teams in Brazil and Sweden, we found that elite players have exceptional cognitive abilities, including improved planning, memory, and decision-making skills. They also possess personality traits like high conscientiousness, extraversion and openness to experience, along with reduced neuroticism and agreeableness. Using artificial intelligence, we identified unique psychological patterns that could help in talent identification and development. These insights can be used to better understand the mental attributes that contribute to success in football and other high-performance fields.
Network Estimation via Source Separation (NESS)
NESS is an innovative framework that I co-developed with Dr Mattia Rosso to derive functional brain networks using linear decomposition techniques. These methods are applicable to various neurophysiological and neuroimaging data and are particularly well-suited to magnetoencephalography (MEG) datasets.
The framework comprises two key methods: BROADband NESS (BROAD-NESS) and FREQuency-resolved NESS (FREQ-NESS).
BROAD-NESS identifies broadband brain networks in event-related designs by applying principal component analysis (PCA) to voxel-level source-reconstructed MEG data. In a musical sequence recognition task, BROAD-NESS revealed the involvement of the auditory cortex in two simultaneous networks. One network, including the medial cingulate gyrus, was linked to auditory processing and sequence monitoring. The other, encompassing hippocampal areas, the inferior temporal cortex, and frontal regions such as the anterior cingulate gyrus, appeared to underpin memory-related processes like prediction matching and prediction error.
FREQ-NESS, in contrast, maps frequency-specific networks using generalised eigendecomposition (GED) on broadband and frequency-specific covariance matrices. It has revealed dynamic brain networks during rest and auditory stimulation, showing how frequency-specific networks spatially reorganise, align with stimulation frequencies, and interact through cross-frequency coupling.
Together, BROAD-NESS and FREQ-NESS provide complementary insights into brain dynamics, uncovering both broadband and frequency-specific networks with high temporal and spatial resolution.
BBC podcast with Leonardo Bonetti: How sonatas help scientists understand the brain
Marco Capogna Young Neuroscientist Prize 2023 assigned to Leonardo Bonetti
Presentations and interviews
In the lab
Oxford fellow
Classical guitarist (in a previous life..)
Videos - Science
Videos - Classical Guitar
Asturias - I. Albeniz
Invocation y danza - J. Rodrigo
Preludio Exarco - L. Bonetti
Fugue BWV 997 - J.S. Bach
Prelude BWV 997 - J.S. Bach
Funebre Exarco - L. Bonetti
Koyunbaba IV - C. Domeniconi
Spanish Dance n.1 - M. De Falla
Finale Exarco - L. Bonetti
Echi - L. Bonetti
Continuum - L. Bonetti
Guitar in a Curved Air - L. Bonetti
Danza - L. Bonetti
Finally... my father proudly embracing a broomstick instead of a gun in his compulsory military service... We must support research and education, not war!
Bonetti, L., Fernandez-Rubio, G., Szabo, S. A., Carlomagno, F., Vuust, P., Kringelbach, M. L., & Brattico, E. (2024). The neural mechanisms of concept formation over time in music. bioRxiv, 2024-11.
Bonetti, L., Fernandez-Rubio, G., Andersen, M. H., Malvaso, C., Carlomagno, F., Testa, C., ... & Rosso, M. (2024). BROADband brain Network Estimation via Source Separation (BROAD-NESS). bioRxiv, 2024-10.
Wehmeyer, L., Baldermann, J.C., Pogosyan, A., Rodriguez Plazas, F., Loehrer, P.A., Bonetti, L., Yassine, S., Zur Muehlen, K., Schueller, T., Kuhn, J. and Visser-Vandewalle, V. (2024). Thalamo-frontal connectivity patterns in Tourette Syndrome: Insights from combined intracranial DBS and EEG recordings. bioRxiv.
Bonetti, L., Risgaard Olsen, E., Carlomagno, F., Serra, E., Szabo, S. A., Klarlund, M., ... & Fernandez-Rubio, G. (2024). Working memory predicts long-term recognition of auditory sequences: Dissociation between confirmed predictions and prediction errors. bioRxiv.
Fernández-Rubio, G., Vuust, P., Kringelbach, M.L., Bonetti, L. (2024). The neurophysiology of healthy and pathological aging: A comprehensive systematic review. bioRxiv.
Campo, F. F., Carlomagno, F., Vuust, P., Haumann, N. T., Bonetti, L., Grube, M., & Brattico, E. (2024). Is auditory prediction related to domain-general cognitive abilities? A neurophysiology study on MMN and cognitive performance scores. A neurophysiology study on MMN and cognitive performance scores. Preprint on SSRN.
Nartallo-Kaluarachchi, R., Bonetti, L., Fernández-Rubio, G., Vuust, P., Deco, G., Kringelbach, M.L., Lambiotte, R. and Goriely, A. (2024). Multilevel irreversibility reveals higher-order organisation of non-equilibrium interactions in human brain dynamics. bioRxiv.
Criscuolo, A., Schwartze, M., Bonetti, L., & Kotz, S. A. (2024). Ageing impacts basic auditory and timing processing. bioRxiv.
Serra, E., Lumaca, M., Brattico, E., Vuust, P., Kringelbach, M. L., & Bonetti, L. (2023). Neurophysiological correlates of short-term recognition of sounds: Insights from magnetoencephalography. bioRxiv.
Quiroga-Martinez, D. R., Rubio, G. F., Bonetti, L., Achyutuni, K. G., Tzovara, A., Knight, R. T., & Vuust, P. (2023). Decoding reveals the neural representation of held and manipulated musical thoughts. bioRxiv.
Costa, M., Vuust, P., Kringelbach, M., Bonetti, L. (2023). Age-related brain mechanisms underlying short-term recognition of musical sequences: An EEG study. bioRxiv.
Brandl, S., Haumann, N.T., Radloff, S., Dähne, S., Bonetti, L., Vuust, P., Brattico, E., Grube, M. (2021). Fourier SPoC: A customised machine-learning analysis pipeline for auditory beat-based entrainment in the MEG. bioRxiv.
Gemma Fernández-Rubio
PhD student, Clinical Medicine
Aarhus University
Elisa Serra
PhD student, Psychiatry
Centre for Eudaimonia and Human Flourishing
University of Oxford
Mathias H. Andersen
Master's student, Psychology
Aarhus University
Ramon Nartallo-Kaluarachchi
PhD student, Mathematics
Centre for Eudaimonia and Human Flourishing
University of Oxford
Chiara Malvaso
Master's student, Physics
University of Bologna
Prof. Morten L. Kringelbach
Professor
Centre for Eudaimonia and Human Flourishing
University of Oxford