Enzo Pasquale Scilingo

Recent advances in physiological sensing technologies enable continuous acquisition of multichannel biosignals for digital health applications. Because physiological interactions arise from nonlinear mechanisms, their characterization requires analytical approaches capable of capturing both individual signal dynamics and cross-signal dependencies, often from short recordings. This study proposes a multilayer visibility graph (VG) framework free of phase-space embedding parameters for analyzing physiological interactions from multichannel biosignals. The approach was evaluated in a simulation study based on unidirectionally coupled Rössler oscillators and in two physiological datasets comprising synchronized cardiovascular recordings from healthy adults and wearable electrodermal activity (EDA) recordings acquired during a Mindfulness-Based Stress Reduction (MBSR) intervention. Average Degree (DA) was used to characterize the topological organization of individual and multiplex visibility graphs, while inter-layer mutual information (I) between degree sequences was introduced to quantify cross-layer dependence. In the simulation study, coupling-related changes in the response dynamics were already detectable from single-layer VGs at short window lengths, whereas inter-layer mutual information captured coupling-dependent changes in cross-layer organization. Feature estimates began to stabilize at approximately 480 samples and approached a plateau at longer window lengths. In the cardiovascular dataset, sex-related differences were primarily observed in configurations involving systolic blood pressure: DA differed in single-layer and multichannel representations, while inter-layer mutual information showed significant sex differences, specifically in the interbeat interval-systolic blood pressure duplex. In the EDA dataset, the proposed descriptors identified changes in trainer–trainee physiological organization across the MBSR intervention. These findings indicate that VG-based descriptors provide complementary information on within-signal topology and inter-layer dependence and can be applied to short multichannel physiological recordings without phase-space reconstruction.

Preterm birth is associated with alterations in early caregiver–infant regulation, with potential consequences for socio-emotional and physiological development. However, the mechanisms through which early interactional experience shapes these processes remain unclear. Here, we tested whether a structured dyadic intervention could modify co-regulatory dynamics across physiological, behavioral, and relational levels. Fifty-four 7-month-old preterm infants and their parents were assigned to either a shared book reading intervention (n = 22) or an active control condition based on a shared building activity (n = 32) and compared with 39 full-term infants. The intervention consisted of an 8-week program of shared book reading, designed to structure parent–infant interaction. Physiological synchrony was assessed at the dyadic level, alongside infants’ autonomic regulation and cardiovascular signal complexity. Behavioral engagement and parental attachment representations were also evaluated. Results showed that mother–infant physiological synchrony emerged selectively within the interactional context trained by the intervention and only in the intervention group. This context-specific synchrony was accompanied by modulation of vagal activity and increased cardiovascular complexity in preterm infants, consistent with enhanced flexibility of autonomic control. At the behavioral and relational levels, intervention infants showed increased initiating joint attention, while parents reported higher secure attachment. These findings support a model of experience-dependent early synchrony, in which repeated dyadic interaction through shared book reading shapes the coupling between interpersonal coordination and individual physiological regulation. By linking synchrony, autonomic flexibility, and social engagement, this study identifies a mechanism through which early caregiving experience can organize developmental trajectories following prematurity.

Whether humans communicate fear through volatile chemical cues has remained unresolved, despite decades of behavioural and neuroimaging evidence. Here, we identify a reproducible molecular signature of acute human fear, bridging the gap between functional evidence and chemical mechanism. Using a low-background sampling approach combined with multiscale chemical analysis and hierarchical modelling, we link high-dimensional axillary volatilomes to a continuous physiological fear index during immersive virtual-reality fear induction. We uncover a coherent chemical fingerprint in which short- and medium-chain fatty acids (acetic acid, butyric acid, caproic acid, and caprylic acid), the caprylic acid precursor octanal, and acetone increase with fear, while skin-derived volatiles (sulcatone, decanal, geranylacetone) and citraconic anhydride decrease. This coordinated shift reveals a fear-associated reorganisation of axillary emissions consistent with sympathetic metabolic mobilisation, involving altered fatty-acid handling, increased ketone-related processes, and suppression of the resident skin volatilome.

Synchrony is a key mechanism that builds up the foundations of human interactions. Quantifying the level of physiological synchronization that occurs during dyadic exchanges is essential to fully comprehend social phenomena. We present a new index to characterize the coupling of complex physiological dynamics: the optimized Multichannel Complexity Index ( op MCI). We validated this approach using synthetic time series of two coupled Hénon Maps, with four different coupling levels in unidirectional and bidirectional manners. We demonstrated that the op MCI method allows to effectively discern between all coupling levels. Then, we applied the op MCI metric on heart rate variability data collected from 37 parent-infant dyads, during shared reading and playing activities, in the framework of the Shared Emotional Reading (SHER) project, with the aim of assessing the effects of early intervention in preterm babies. Two groups presented preterm infants: an intervention group, who participated in a two-month shared reading program, and a control group, who practiced shared play activities. A full-term group provided additional control data. The op MCI values were significantly higher for the intervention dyads with respect to the other groups during the shared reading task, showing that an early reading intervention program could increase parent-infant synchrony in preterm babies. Graphical Highlights: Interpersonal physiological synchrony is a fundamental mechanism for the evolutionary path An ad-hoc methodological framework to assess nonlinear cardiac coupling is lacking The op MCI has been shown to be a reliable tool for assessing the degree of synchrony Early shared book reading may increase parent-infant synchrony in a clinical setting

The ability to establish and maintain relationships with users is crucial for Interactive Conversational Agents (ICAs). This involves remembering information from previous interactions and utilizing it in future conversations-key behaviors that build rapport and demonstrate interest in the user, which are essential for sustained engagement. Current conversational agents typically rely on context from recent conversational turns. To enhance interaction quality, agents should be equipped with a Long-Term Memory (LTM) that encompasses entire past conversations. The state-of-the-art approach to incorporating LTM in ICAs involves saving all previous turns in an external storage system (e.g., a plain database). Despite achieving considerable results, this method is not biologically inspired and diverges from how human memory works-where only essential information is stored and reconstructed generatively using semantic memory. We propose a hierarchical memory system inspired by human memory stratification. It dynamically updates and retrieves contextually relevant episodic and declarative memories to support semantically enriched interactions. A significant innovation in our approach is the use of a Large Language Model (LLM) Agent to determine what structured information should be stored during conversations. Our preliminary results suggest that our approach maintains competitive performance compared to existing methods while introducing novel features such as graph-based persona representation and a reduced need for storing entire past dialogues.

Large Language Models (LLMs) now possess function calling capabilities, enabling them to interact with external tools and APIs. Within cognitive architectures for social robotics, this provides a robust mechanism for an LLM to orchestrate a set of discrete functions-conceptually echoing the brain's functional specificity-to manage operations such as visual perception, auditory processing, speech generation, and memory access, among others. However, LLMs exhibit varying propensities and strategies when utilizing these function calls. This paper presents a comparative evaluation of four LLMs, differing significantly in parameter scale and origin: Llama3 70b, Gemma2 9b, Mixtral 8x7b, and Phi3 mini 3.8b, each acting as the orchestrator in such an architecture. Our testbed involved a human-participant study (N=20) where individuals engaged in ambiguous social scenarios, each interacting with the architecture driven by the different LLMs (4 trials per participant). Results revealed statistically significant differences in the frequency of Look (F=13.62, p<0.001), Talk (F=9.29, p<0.001), and Hear (F=10.34, p<0.001) calls across LLMs. Notably, Llama3 70b made significantly more ‘Look’ calls (M=3.45), a behavior that corresponded with strong user preference (18/20), suggesting its interaction style was perceived as more natural and contextually aware. Mixtral 8x7b, in contrast, favored ‘Talk’ (M=11.05) and ‘Hear’ (M=11.15) calls. These findings demonstrate that analyzing function call patterns offers a quantitative lens to understand and compare the interaction strategies of different LLMs in orchestrating robotic behavior.

In Affective Computing, a key challenge lies in reliably linking subjective emotional experiences with objective physiological markers. This preliminary study addresses the issue of reproducibility by identifying physiological features from cardiovascular and electrodermal signals that are associated with continuous self-reports of arousal levels. Using the Continuously Annotated Signal of Emotion dataset, we analyzed 164 features extracted from cardiac and electrodermal signals of 30 participants exposed to short emotion-evoking videos. Feature selection was performed using the Terminating-Random Experiments (T-Rex) method, which performs variable selection systematically controlling a user-defined target False Discovery Rate. Remarkably, among all candidate features, only two electrodermal-derived features exhibited reproducible and statistically significant associations with arousal, achieving a 100% confirmation rate. These results highlight the necessity of rigorous reproducibility assessments in physiological features selection, an aspect often overlooked in Affective Computing. Our approach is particularly promising for applications in safety-critical environments requiring trustworthy and reliable white box models, such as mental disorder recognition and human-robot interaction systems.