Jing Xia is an Assistant Researcher at the Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, and a Ph.D. candidate at South China Normal University with joint doctoral training at Ruhr University Bochum, Germany. She has been selected as a Shenzhen High-level Talent, Pengcheng Outstanding Talent, and Nanshan District Leading Talent. She has long focused on audio-visual channel cognition and music therapy for emotional disorders, and is committed to developing intelligent diagnosis systems for brain diseases based on electrophysiological signals, conducting research on brain diseases such as depression.

As the primary author, she has published 14 papers on cognitive neuroscience and brain diseases in international journals such as Cell Reports, NeuroImage, and Cortex, and top conferences such as ICASSP and IEEE EMBC. She has applied for 3 patents related to EEG signal processing and brain disease diagnosis, with 1 granted. She has hosted 4 projects including the Guangdong Provincial Natural Science Foundation, China Postdoctoral Science Foundation, and Shenzhen General Program, and participated in 9 projects including the National Key R&D Program, NSFC General Program, Guangdong Provincial Key Program, and Shenzhen Key Program.

Music can induce emotional experiences through physical properties such as rhythm, mode, and amplitude envelope, but how it forms quantifiable and controllable neural entrainment mechanisms in deep reward circuits still lacks direct evidence. This talk focuses on musical acoustic features and deep neural dynamics, introducing a music brain-computer interface research framework for music-modulated implicit emotion.

First, the study utilizes the high spatiotemporal resolution of stereo-electroencephalography (SEEG) to investigate the effects of different musical physical properties on time-frequency activity, cross-frequency coupling, and network connectivity of dopamine reward circuit nodes such as the nucleus accumbens, striatum, and medial prefrontal cortex, establishing a nonlinear mapping between musical physical parameters and neuroelectrophysiological responses. Then, in epilepsy patients comorbid with depression, intervention is carried out using individually most sensitive musical parameters, combined with implicit emotion tests to extract neural responses of deep dopamine circuits.

On this basis, the study introduces a Surrogate Brain architecture combining neural mass models with data-driven learning to learn cross-modal mappings between EEG and SEEG. The system can infer deep reward states in real time relying solely on scalp EEG. The talk will discuss the key challenges of scarce clinical data at the intersection of music information processing, affective computing, neuroengineering, and digital therapeutics. The research aims to provide verifiable neural mechanisms and translatable non-pharmacological digital therapeutic pathways for music intervention in emotional disorders.