ABSTRACT NUMBER: ESOC2026A2072 PRELIMINARY CONNECTOME-BASED INSIGHTS INTO EMOTION RECOGNITION AFTER SUCCESSFUL THROMBECTOMY

Abstract Background and aims Emotion recognition is a core component of social cognition and essential for interpersonal functioning. However, despite good functional outcomes after endovascular thrombectomy (EVT), subtle neurocognitive deficits may persist, with impairments in emotion recognition representing a common yet under-recognized consequence associated with reduced psychosocial functioning. This study aims to characterize emotion recognition performance after EVT and relate behavioral deficits to network-level structural disconnection using a connectome-based approach. Methods This study is part of the NESCAT study (Neuropsychiatric Sequelae and Cognitive Deficits after Stroke with Successful Thrombectomy and Good Functional Outcome), a prospective cohort of EVT-treated ischemic stroke patients with neuropsychological follow-up at three months. Emotion recognition was assessed using a facial emotion recognition task. Stroke lesions were segmented semi-automatically and used to quantify structural disconnection with a connectome-based approach. Associations between disconnection measures and emotion recognition performance were examined using multivariable regression. Results Across emotions, accuracy varied substantially, with the lowest performance observed for fear recognition. Exploratory regression analyses demonstrated a significant age-related decline in emotion recognition accuracy. While hemispheric lesion side contributed to emotion recognition performance, a large proportion of interindividual variability remained unexplained, motivating the use of more fine-grained, network-level analyses. Ongoing connectome-based analyses will determine whether distributed structural disconnection within socio-emotional networks explains these deficits beyond demographic and lesion-level factors. Conclusions Preliminary behavioral findings indicate emotion-specific vulnerabilities, particularly for fear recognition, despite good functional outcomes after EVT. This study aims to extend lesion-level analyses toward a connectome-based understanding of subtle emotion recognition deficits after stroke. Conflict of interest Dr. Naomi Giesers: nothing to disclose