Biology subjects

Oddo, C. M.

Publications and source records attributed to Oddo, C. M..

3 recordsLinked to original sources

bioRxiv2026

Resting state EEG microstates: is a new FORM all you need?

Introduction: Before asking which microstate an EEG sample belongs to, there is a simpler question: can it look like a microstate at all? Yet a winning correlation mixes three different properties: how microstate-like the scalp field is, where it lies within that geometry, and how well the available templates represent that direction. We asked whether a fixed geometry, independent of any microstate solution, could explain common structure of microstate maps, constrain how strongly EEG patterns can match them, and describe that structure in interpretable terms. Methods: We introduce First-Order spherical haRMonic (FORM), projecting normalized scalp topographies onto a fixed three-dimensional subspace defined only by sensor geometry. Direction describes field geometry; projection norm, {rho}, measures FORM conformity and the maximum correlation attainable by any FORM topography. We tested simulations, literature templates, and 374 recordings from 187 LEMON participants. Results: FORM captured canonical microstate geometry despite being defined a priori of a microstate solution: FORM explained energy, {rho} squared, was 0.990-0.998 for five meta-microstates and 0.966-0.986 for pooled LEMON maps. Across 313 literature templates, FORM angular distances preserved their independently derived organization (Spearman {rho}=0.987). In continuous EEG, FORM conformity explained a median 74.4% of within-subject variance in winning-template correlation, versus 21.7% for GFP. As independently estimated dictionaries increased from k=1 to k=12, the median fitted fraction of the FORM ceiling rose from 0.613 to 0.970. Discussion: FORM places an a priori continuous geometry beneath discrete microstate labels. It separates properties intrinsic to the scalp field - whether it is microstate-like and where it lies within that geometry - from how well the chosen template dictionary represents it. FORM therefore provides an interpretable, clustering-independent reference for assignment strength, template comparison and continuous microstate dynamics.

neuroscience↗

bioRxiv2026

Spatial localisation of touch on a robotic limb in the absence of direct haptic feedback

Artificial limbs typically lack somatosensory receptors, limiting users access to haptic feedback about movement outcomes. However, indirect tactile signals generated during object interaction are transmitted to the body-limb interface and may serve as a haptic feedback source. This study investigates whether such indirect tactile signals convey distinct information that can support localisation of touch on an artificial body part. Using an additional robotic thumb (The Third Thumb, Dani Clode Design), we measured tactile signals at the interface between the Third Thumb and the hand following stimulation from four vibration motors positioned along the Thumb. We then assessed whether participants could discriminate these signals and integrate them with different Thumb postures to spatially localise touch. Measurements showed that each stimulation site produced distinct tactile signatures at the interface. Behaviourally, participants localised touch at above-chance levels. Representational similarity analyses further revealed that performance was best explained by the spatial positions of stimuli, rather than by limb-based coordinates or indirect signal patterns alone, indicating flexible mapping of indirect touch signals onto the Third Thumb posture. Together, these findings demonstrate that indirect tactile signals transmitted through an artificial limb can be discriminated and flexibly remapped to support spatial localisation of touch.

neuroscience↗

bioRxiv2021

Euler method can outperform more complex ODE solvers in the numerical implementation of the Izhikevich artificial Spiking Neuron Model given the allocated FLOPS

The Izhikevich artificial spiking neuron model is among the most employed models in neuromorphic engineering and computational neuroscience, due to the affordable computational effort to discretize it and its biological plausibility. It has been adopted also for applications with limited computational resources in embedded systems. It is important therefore to realize a compromise between error and computational expense to solve numerically the models equations. Here we investigate the effects of discretization and we study the solver that realizes the best compromise between accuracy and computational cost, given an available amount of Floating Point Operations per Second (FLOPS). We considered three fixed-step solvers for Ordinary Differential Equations (ODE), commonly used in computational neuroscience: Euler method, the Runge-Kutta 2 method and the Runge-Kutta 4 method. To quantify the error produced by the solvers, we used the Victor Purpura spike train Distance from an ideal solution of the ODE. Counterintuitively, we found that simple methods such as Euler and Runge Kutta 2 can outperform more complex ones (i.e. Runge Kutta 4) in the numerical solution of the Izhikevich model if the same FLOPS are allocated in the comparison. Moreover, we quantified the neuron rest time (with input under threshold resulting in no output spikes) necessary for the numerical solution to converge to the ideal solution and therefore to cancel the error accumulated during the spike train; in this analysis we found that the required rest time is independent from the firing rate and the spike train duration. Our results can generalize in a straightforward manner to other spiking neuron models and provide a systematic analysis of fixed step neural ODE solvers towards an accuracy-computational cost tradeoff.

bioengineering↗