ABC Seminar: Human brain imaging
Welcome to the ABC Seminar! This seminar series is open for everyone. The talk will take place in the Room U141 (U3) in Otakaari 1. After the talk, coffee and pulla will be served.
The event will be also streamed via Zoom at: https://aalto.zoom.us/j/61645942470
Speaker: Philippe Ciuciu (CEA/NeuroSpin & Inria Parietal)
Title: Accelerated brain MR Imaging: From shorter data acquisition to faster image reconstruction
Magnetic Resonance Imaging is a widely used neuroimaging technique as it can probe brain tissues, their structure and provide insights on the functional organization as well as the layout of brain vessels. However, MRI relies on an inherently slow imaging process.
Reducing acquisition time has been a major challenge in high-resolution MRI and has been successfully addressed by Compressed Sensing (CS) theory. Nevertheless, most of the Fourier encoding schemes under-sample existing k-space trajectories which does not adequately encode all the information necessary. Recently, my team has addressed this issue by proposing the Spreading Projection Algorithm for Rapid K-space sampLING (SPARKLING) for 2D/3D non-Cartesian T2* and susceptibility weighted imaging at 3 and 7Tesla (T). In the first half of my presentation, I will present these advancements, interesting clinical applications and how we have adapted this approach to address high-resolution functional MRI at 7T.
However, CS approaches suffer from a slow image reconstruction process. To counteract this delay and improve image quality, deep learning has been used since 2016. In the second half of this talk, I will expose our own deep-learning architecture, called XPDNet (Primal Dual Network where X plays the role of a magic card), that has been ranked second in the 2020 brain fastMRI challenge (1.5 and 3T data). I will illustrate XPDNet’s transfer learning capacity on 7T NeuroSpin T2 images. Lastly, I will share how we have further improved this approach to handle 3D non-Cartesian imaging settings associated with the SPARKLING encoding scheme.
Dr. Philippe Ciuciu obtained his PhD in electrical engineering in 2000 and his Habilitation to Conduct Research degree in 2008 both from the University of Paris-Sud (Orsay, France). The early stages of his career began at the CEA (The French Alternative Energies and Atomic Energy Commission), then in 2012-13 he was invited by the Department of Applied Mathematics at the University of Toulouse as a guest professor. Dr. Ciuciu is now a CEA Research Director at NeuroSpin (CEA), the largest ultra-high field MRI center in France dedicated to cognitive and clinical neuroscience research. He has a joint appointment with Inria, the French institute of computer Science and automatic control where he has led, since 2018, the Compressed Sensing team in the Inria-CEA Parietal unit located at NeuroSpin. In Jan. 2022, due to his expertise in neuroimaging techniques, brain data analysis, and machine learning, he will co-lead the new joint Inria-CEA unit MIND, (Models and Inference for Neuroimaging Data), a cutting edge research unit comprised of 40 researchers and staff.
His inter-disciplinary research interests range from methodological developments in accelerated MRI to cuttingedge signal processing tools for the analysis of functional brain data (fMRI, M/EEG) with applications in cognitive and clinical neuroscience (neurodevelopment and neurodegeneration). His work has led to over 200 peer-reviewed publications including 3 MRI related patents and more than 55 articles in international journals.
As an IEEE Senior Member, Dr. Ciuciu has represented the IEEE Signal Processing Society in the International Symposium on Biomedical Imaging for the 2019-2022 period. He has driven the topic related to Artificial Intelligence in MRI in the steering committee of the virtual 2021 ESMRMB conference. As of 2019, he holds a position as Senior Area Editor for the IEEE open Journal of Signal Processing and that of Vice Chair for the Biomedical Image and Signal Analytics (BISA) technical committee of the EURASIP society. In 2020, he was appointed as Associate Editor to Frontiers in Neuroscience, section: Brain imaging methods.
Speaker: Matias Palva (Aalto University)
Title: New vistas for brain criticality
Understanding the determinants of neuronal dynamics in the brain is central for understanding how cognitive and mental functions arise in complex and intricate neuronal networks. “Brain criticality” is a framework for characterizing dynamics of collective neuronal activity and understanding the factors regulating this collectivity. The key tenets here are that (i) the brains are likely to operate near a critical point at the phase transition between disorder and order, and (ii) operating in such a regime yields several functional benefits and may be crucial for healthy brain functioning.
Several lines of experimental evidence, both in animal models and human electrophysiological recordings, support the notion of brain criticality. In healthy brains, critical brain dynamics predicts both behavioral dynamics and cognitive performance, e.g., in tasks demanding cognitive flexibility. Conversely aberrant critical dynamics characterize many brain diseases, predict symptoms, and may even play a causal role in the underlying pathogenesis.
Recent findings, however, call upon an extension of the classical statistical-physics inspired brain criticality framework. First, theoretical studies have shown that heterogeneity, and especially modularity, in the interaction networks stretches the classical critical point into an extended regime. Second, both theoretical and experimental studies suggest that brain dynamics involve bistability and bistable critical dynamics. The implications of these phenomena/mechanisms on brain criticality will be discussed.
Matias Palva group studies the systems-level neuronal mechanisms of emergent neuronal and behavioral dynamics.
Spontaneous brain activity fluctuates in time scales spanning at least across five orders of magnitude. These fluctuations have also been observed on all studied spatial scales and they are statistically governed by spatio-temporal power-laws.
Such a scale-free organization at a macroscopic level is, however, contrasted by salient scale-specific neuronal activities - neuronal oscillations. Our research addresses the functional significance of scale-free and scale-specific brain dynamics in human sensory perception, cognitive performance, and motor output.
We have developed methods for MEG/EEG source reconstruction, optimized cortical parcellations, and quantification of neuronal/behavioral scaling-laws as well as for the mapping of dynamic neuronal interaction networks from invasive and non-invasive electrophysiological recordings of human brain activity. We are also in the process of translating our data management, analysis, and visualization platform into a more easily shareable python package.
Our three main research lines are 1. Assessing the functional roles of brain criticality and connectivity in human cognition by using MEG/EEG and SEEG based connectomes of neuronal couplings and "dynomes" of spatio-temporal dynamics. We are also performing simulations of brain dynamics and utilize several lines of interventional approaches, from electric and magnetic brain stimulation to cognitive training. 2. Identifying the roles of dysconnectivity and dysdynamics in mental disorders such as depression, anxiety, ADHD and schizophrenia, with the major depressive disorder being our main research focus. 3. Developing neuroplasticity-recruiting cognitive training methods for targeted alterations of cortical connectivity and dynamics.