Congratulations to Jingxu Liu, a Master's Student in Professor Duoqian Miao's Research Group, on the Acceptance of His Paper “Visual-affective decoding enhances neurofunctional objective and generalizable signatures in unseen adolescent depression diagnosis” by npj Digital Medicine
Recently, the paper "Visual-affective decoding enhances neurofunctional objective and generalizable signatures in unseen adolescent depression diagnosis" by master student Jingxu Liu, under the guidance of Prof. Duoqian Miao, has been accepted by npj Digital Medicine, a leading journal at the intersection of digital medicine and AI in healthcare. The study introduces a Visual–Affective Decoding (VAD) framework using naturalistic movie-watching fMRI data and a normative dual-pathway anomaly detection model. Experimental results show strong cross-site generalizability and reveal interpretable neurofunctional signatures, offering effective support for adolescent MDD diagnosis and biomarker development.
近日，实验室苗夺谦教授指导的2024级硕士研究生刘景旭的论文"Visual-affective decoding enhances neurofunctional objective and generalizable signatures in unseen adolescent depression diagnosis"被国际顶级期刊npj Digital Medicine正式录用。npj Digital Medicine是数字医学与人工智能医疗交叉领域公认的顶级期刊，长期位列中科院一区 Top 期刊，同时作为 Nature 旗下npj 系列成长极快具有极高声誉的学术期刊，在国际学术界具有重要影响力。
针对青少年重度抑郁症缺乏客观且具泛化能力的神经功能特征、现有影像检测方法缺乏生态效度等问题，该研究提出了“视觉-情感解码”（VAD）框架。该方法整合了来自北京安定医院提供的自然刺激电影观看期间的全脑 fMRI 数据，并构建了基于健康对照组的双解码通道异常检测模型，通过偏离规范脑功能的程度来识别抑郁症。实验结果表明，该方法具备良好的跨中心泛化能力，并能揭示可解释的神经功能特征，为青少年抑郁症的诊断及临床生物标志物的开发提供了有效支撑。
