Christine Leilani Skjegstad

• Beinnes, Gard Kvammen; Skjegstad, Christine Leilani & Frühholz, Sascha (2025). Differential neural decoding of alarm and avoidance information from vocal alarm calls in humans. Communications Biology. 8(1). doi: 10.1038/s42003-025-08248-9. Full text in Research Archive Show summary

  Many animals and humans use scream calls to signal imminent threats, which typically evoke alarm and escape responses in recipients. Compared to animals, human scream calls are more diversified with varying levels of alarm signaling. Certain low-alarm human screams thus might not elicit full neural alarm cascades and threat avoidance actions. Here we investigated the neural circuits for decoding alarm and avoid-approach information from positive and negative scream calls in humans. Alarm and avoid-approach decisions showed minimal neural overlap, with alarm decisions recruiting neural cascades of sensory-affective decoding for action preparation, whereas avoid-approach judgments recruited neural systems for spatio-affective decoding for social decision-making. Furthermore, decision patterns revealed both an alarm quotient and an avoid quotient for risk arbitration, which linked the likelihood of an urgent response (alarm, avoid) to a slowing of choosing a potential safe option in response to scream calls. While the avoid quotient positively predicted neural activity in a broad cortico-limbic network, the alarm quotient predicted neural suppression with increasing alarm levels, especially in the amygdala as part of a presumed limbic alarm system. This points to a critical involvement of the amygdala at neural levels of choice arbitrations rather than in threat evaluations signaled by screams.

• Skjegstad, Christine Leilani; Trevor, Caitlyn; Swanborough, Huw; Roswandowitz, Claudia; Mokros, Andreas & Habermeyer, Elmar [Show all 7 contributors for this article] (2022). Psychopathic and autistic traits differentially influence the neural mechanisms of social cognition from communication signals. Translational Psychiatry. 12(1). doi: 10.1038/s41398-022-02260-x. Full text in Research Archive Show summary

  Abstract Psychopathy is associated with severe deviations in social behavior and cognition. While previous research described such cognitive and neural alterations in the processing of rather specific social information from human expressions, some open questions remain concerning central and differential neurocognitive deficits underlying psychopathic behavior. Here we investigated three rather unexplored factors to explain these deficits, first, by assessing psychopathy subtypes in social cognition, second, by investigating the discrimination of social communication sounds (speech, non-speech) from other non-social sounds, and third, by determining the neural overlap in social cognition impairments with autistic traits, given potential common deficits in the processing of communicative voice signals. The study was exploratory with a focus on how psychopathic and autistic traits differentially influence the function of social cognitive and affective brain networks in response to social voice stimuli. We used a parametric data analysis approach from a sample of 113 participants (47 male, 66 female) with ages ranging between 18 and 40 years (mean 25.59, SD 4.79). Our data revealed four important findings. First, we found a phenotypical overlap between secondary but not primary psychopathy with autistic traits. Second, primary psychopathy showed various neural deficits in neural voice processing nodes (speech, non-speech voices) and in brain systems for social cognition (mirroring, mentalizing, empathy, emotional contagion). Primary psychopathy also showed deficits in the basal ganglia (BG) system that seems specific to the social decoding of communicative voice signals. Third, neural deviations in secondary psychopathy were restricted to social mirroring and mentalizing impairments, but with additional and so far undescribed deficits at the level of auditory sensory processing, potentially concerning deficits in ventral auditory stream mechanisms (auditory object identification). Fourth, high autistic traits also revealed neural deviations in sensory cortices, but rather in the dorsal auditory processing streams (communicative context encoding). Taken together, social cognition of voice signals shows considerable deviations in psychopathy, with differential and newly described deficits in the BG system in primary psychopathy and at the neural level of sensory processing in secondary psychopathy. These deficits seem especially triggered during the social cognition from vocal communication signals.

View all works in NVA

• Skjegstad, Christine Leilani; Abrahamsson, Liv Merve Akca & Frühholz, Sascha (2025). Social traits influence behavioral and neural responses to emotional AI and human voices. doi: https:/www.abstractsonline.com/pp8/#%21%2F21171%2Fpresentation%2F23566. Full text in Research Archive Show summary

  Individual differences tint our perception of reality. Our previous work showed certain personality traits influence how the brain responds to human voices. Recently, we found the brain responds differently to voices generated by artificial intelligence (AI) compared to human voices, despite challenges in perceptual discrimination. This study investigates whether social traits further influence differential neural processing of AI and human voices. 61 participants (40 females, mean age = 31.93, SD = 7.75) listened to vocal emotional pseudo-speech expressions (neutral, angry, fear, happy, pleasure) from both AI-generated and human sources. The protocol included an fMRI study with a classification task, followed by a perceptual task, as well as personality trait assessments. Traits related to social and emotional cognition (Callous-Unemotional [CU], Alexithymia [ALX], Affective Empathy [AE], and Social Information Processing [SIP]) were selected. Correlations between the traits and outcome measures were performed (p <0.05, FDR corrected). ALX correlated with a tendency to classify voices as AI (r = -0.26) while SIP correlated with a tendency to classify voices as human (r = 0.3). AE and SIP were positively correlated with perceived naturalness of human voices (r = 0.32 and r = 0.27). Conversely, ALX and CU were negatively associated with perceived naturalness of human voices (r = -0.28 and r = -0.28). CU negatively correlated with perceived naturalness of AI voices (r = -0.28), whereas AE positively correlated with it (r = 0.29). Our fMRI multiple regression analysis included the traits with age, sex, and handedness as covariates of no interest (p < 0.005, k > 21). Our findings reveal distinct neural patterns: in response to AI voices (versus human), AE was associated with increased activity in the left angular gyrus, left inferior frontal gyrus-triangular part, left hippocampus, right nucleus accumbens and bilateral amygdalae. SIP was associated with increased activity in the right dorsolateral superior frontal gyrus. In response to human voices (versus AI), AE was associated with increased activity in the left precentral gyrus, while CU was associated with increased activity in the bilateral precentral gyri. SIP was associated with increased activity in the right cuneus, right superior temporal gyrus, right fusiform gyrus, left middle cingulate and paracingulate gyri, left insula, and bilateral middle occipital gyri. When processing AI voices, AE may relate to emotional engagement and SIP to more cognitive effort. In the context of human voices, AE and CU may differentially engage the mirror neuron system, while SIP may engage multimodal processing.

• Skjegstad, Christine Leilani & Frühholz, Sascha (2024). Neural Dynamics of Processing Natural and Digital Emotional Vocalizations. Full text in Research Archive Show summary

  AI-generated voices have a high level of auditory quality and resemble human voices, such that they might mimic human emotions to a high degree. In our study, we investigated how the human brain distinguishes between emotional natural (human) and synthetic (AI-generated) voices. 35 participants (23 females, mean age = 32.23, SD = 7.66) listened to vocal pseudo-speech emotional expressions (neutral, angry, fear, happy, pleasure) from both human and AI-generated sources. The experiment consisted of three parts: an fMRI study with a classification task to identify the voices as synthetic or natural, a perceptual task where participants rated characteristics of the voice stimuli, and an assessment of personality traits. Neutral AI voices were identified with 74.9% accuracy compared to only 23% accuracy for neutral human voices. Happy human voices were correctly identified 77% of the time while happy AI voices were identified with 34.5% accuracy. In general, human voices were rated as more natural, trustworthy, and authentic, especially the happy and pleasure expressions. Neuroimaging results revealed significant differences in brain activity in response to the AI and human voices. AI voices activated the right anterior midcingulate cortex, right dorsolateral prefrontal cortex and left thalamus, which may indicate increased vigilance and cognitive regulation. In contrast, human voices elicited stronger responses in the right hippocampus as well as regions associated with emotional processing and empathy such as the right inferior frontal gyrus, anterior cingulate cortex and angular gyrus. Human voices may elicit a sense of relatedness and AI voices may elicit heightened alertness.

• Skjegstad, Christine Leilani & Noble, Kerry (2024). FENS Forum 2024 media programme – Our brains respond differently to human and AI-generated speech, but we still struggle to tell them apart. [Internet]. FENS. Full text in Research Archive

View all works in NVA