Research areas in which I am invested are orthopedic surgery, rheumatology, and neuroscience. I have been working with Pr Christian Roux (Clinical Research Center) and Dr Marc-Olivier Gauci (Institut Universitaire Locomoteur et Sport). I contribute to the design, methodology, submission and development of clinical research protocols and research and innovation projects (ANR, FHU, Tiers Lieux Experimentation) and ensured contacts with startups and pharmaceutical companies in the corresponding domains.
Main topic - Orthopaedic Surgery & Imaging advances, and Rhematologic diseases & pharmaceutical solutions & virtual reality.
Context - Orthopedic Surgery and Traumatology (deep-tech AI-based innovative oriented projects), Rheumatology & Neuroscience (Transdisciplinary solutions to decrease pain and improve patient quality of life)
My main research domain is computational functional MRI (fMRI) for Brain Connectomics, that is inserted in the context of the CoBCoM ERC advanced grant project. The main objective of my project focused on methods to recover the brain functional activations, their dynamics and connectivity, starting from corrupted fMRI data. The project has been supervised by Dr. Rachid Deriche and the co-supervised by Dr. Samuel Deslauriers-Gauthier.
Main topic - Brain functional activation recovery.
Context - Neuroimaging modalities give indirect and degraded measures of the brain activity or structure, meaning that the acquired signals are corrupted by noise, and have to be interpreted using sophisticated analysis methods which allow to extract meaningful information.
Problem and state of the art - Traditional techniques for measuring the brain function are based on experimental paradigms, where the subject is asked to perform a task to measure the difference between the rest condition, representing a baseline, and the task condition. Nevertheless, an experimental setup is not suitable for those patients whose conditions do not allow them to perform tasks. Interestingly, the development of new techniques in the field of fMRI, i.e. the resting-state fMRI (rs-fMRI), provide signals that may give insights into the brain function in the absence of stimuli, when the subject is at rest, and is not required to perform any task. This has emphasized the need to recover the underlying neural activations from fMRI signals in the absence of an experimental paradigm.
Objective - Provide approaches for the recovery of functional brain activations, without a priori information on their timing, duration and location.
Strategy - Design, implementation and validation of:
Temporal regularized deconvolution of the blood-oxygen-level-depenent (BOLD) signal using the Least Absolute Shrinkage and Selection Operator (LASSO) model, solved by means of the Least-Angle Regression (LARS) algorithm.
The Paradigm-Free fMRI (PF-fMRI) is a method to blindly recover brain functional activations from resting-state fMRI data, where the subject is at rest and no a priori information on the location and timing of the underlying activation is given.
Both approaches allowed us to recover the underlying neurons activations and their dynamics.
Check out the following video for a more detailed explanation of the PF-fMRI.
The PF-fMRI is presented in two journal papers:
Costantini I., Deriche R., Deslauriers-Gauthier S., “A Paradigm Free Regularization Approach to Recover Brain Activation from fMRI Data”, manuscript in preparation for submission.
Deslauriers-Gauthier S., Costantini I., Deriche R.,”Non-invasive inference of information flow using diffusion MRI, functional MRI and MEG”, Journal of Neural Engineering, 2020.
The PF-fMRI was presented in two conference talks:
Talk at ISMRM 28th Annual meeting & exhibition 2020, (Virtual) for the abstract (1224, https://www.ismrm.org/20/program_files/O77.htm)
Talk at ISMRM 27th Annual meeting & exhibition 2019, (Montreal) for the abstract (910, https://www.ismrm.org/19/program_files/PP26.htm)
and it earned the Magna Cum Laude award by the ISMRM Society (https://www.ismrm.org/20/program_files/O69.htm) in 2020.
M. Frigo, I. Costantini, R. Deriche, and S. Deslauriers-Gauthier, (2018, September). Resolving the crossing/kissing fiber ambiguity using Functionally Informed COMMIT. In International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI). Springer, Cham.
M. Frigo, GG. Diez, I. Costantini, A. Daducci, D. Wassermann, R. Deriche, and S. Deslauriers-Gauthier, Reducing false positive connection in tract function filtering, OHMB 2018, Singapore.
Main topic - Brain connectivity.
Context - The Default Mode Network (DMN) is the most studied of the resting state networks (RSNs) detected in rs-fMRI, since it is known to be deactivated by tasks and is hypothesized to play an important co-ordination role in cognition.
Problem and state of the art - One of the most widespread methods is the independent component analysis (ICA), that allows to find patterns of activation common to a heterogeneous group of subjects in absence of a priori hypotheses. However, this approach does not allow a within-network study of functional connectivity, thus overlooking important information in order to understand age and gender effects on the physiological changes of functional connectivity.
Objective - Provide a method to explain evolution and changes in functional connectivity of the DMN sub-networks in regard to age and gender, using an innovative method as a complement to ICA.
Strategy - Design, implementation and validation of a innovatice approach that uses ICA to extract RSNs common to the analyzed group of subjects and a clustering algorithm to further parcel the DMN into its anatomically different sub-networks.
The figures below shows some results.
RSNs obtained with state of the art approach. (IC: independent component)
Clusters obtained with the proposed mixed approach. (CL: cluster)
Functional connectivity between DMN regions was found to increase from childhood to adulthood and then slowly decrease with ageing. Also, females were found to develop the left hemisphere FC earlier than males and to lose left FC sooner than males.
The research was the outcome of a collaboration between the Don Carlo Gnocchi Foundation and Politecnico di Milano .
Journal papers
Quatto P., Margaritella N., Costantini I., Baglio F., Garegnani M., Nembi R., & Pugnetti L., “Brain networks construction using Bayes FDR and Average Power Function”, Statistical Methods in Medical Research, 2019.
Facco E., Mendozzi L., Bona A., Motta A., Garegnani M., Costantini I., Dipasquale O., Cecconi P., Menotti R., Coscioli E., Lipari S., “Dissociative identity as a continuum from healthy mind to psychiatric disorders: epistemological and neurophenomenological implications approached through hypnosis”, Medical Hypotheses, 2019.
Finotelli P., Dipasquale O., Costantini I., Pini A., Baglio F., Baselli G., Duilio P., & Cercignani M., “Exploring resting-state functional connectivity invariants across lifespan in healthy people by means of a recently proposed graph theoretical model”, PloS one, 2018.
Marchetti A., Baglio F., Costantini I., Dipasquale O., Savazzi F., Nemni R., Sangiuliano Intra F., Tagliabue S., Valle A., Massaro D. & Castelli I., “Theory of mind and the whole brain functional connectivity: behavioral and neural evidences with the Amsterdam Resting Questionnaire”, Frontiers in Psychology, 2015.
Conference abstracts
Castagna A., Sciume L., Costantini I., Ramella M., Montesano A., Baglio F., “Long lasting cerebral functional connectivity changes of a rehabilitation perceptive integrated approach associated with botulinum toxin A in cervical dystonia”, Journal of Neurotransmission, 2016.
Laganà M., Baglio F., Pelizzari L., Dipasquale O., Costantini I., Baselli G., Bergsland N., Lecconi P., Clerici M., Haacke M., Mendozzi L., Nembi R., “Combined study of neurodegeneration, cerebrovascular reactivity and venous drainage ”, at ISNVD 2016 Annual Meeting, 2016.
Laganà M., Baglio F., Pelizzari L., Dipasquale O., Costantini I., Baselli G., Bergsland N., Lecconi P., Clerici M., Haacke M., Mendozzi L., & Nemni R., “Brain perfusion and venous drainage in Multiple Sclerosis: a multimodal approach.”, at OHBM 2016 Annual Meeting, 2016.
Pelizzari L., Scaccianoce E., Dipasquale O., Costantini I., Baglio F, & Baselli G.,”Correlation of Brain Structural and Functional Connectivity Indexes“, 37th Anual International Conference of the IEEE Engineering in Medicine and Biology Society, 2015.
Baselli G., Bergsland N., Costantini I., Dipasquale O., Scaccianoce E., Laganà M., Pelizzari L., Clerici M., Baglio F., “Integrating structural and functional brain connectivity image, signal, and data processing problems”, at 2015 AEIT International Annual Conference, 2015.
Baselli G., Bergsland N., Costantini I., Dipasquale O., Scaccianoce E., Laganà M., Pelizzari L., Clerici M., & Baglio F., “Analisi di connettività encefalica strutturale e funzionale”, AEIT (ICT), 2017.