NCS-FO: SOUND: Understanding the Functional Neural Dynamics Underpinning Auditory Processing Dysfunctions through a Multiscale Recording-Stimulation Framework

听觉加工功能障碍（APD）是包括精神分裂症在内的许多类型精神病的共同特征，并与多种核心症状相关，包括听觉言语幻觉（即听到声音）。尽管APD的患病率很高，影响了高达80%的精神病人群，但药物治疗是无效的：70%的患者要么有不良副作用，要么在治疗后仍持续出现症状。非药物治疗策略，如神经调节（有针对性的神经刺激），将满足重要的医疗需求。尽管神经调节最近已成为一系列神经心理疾病的合理治疗工具，但对APD背后的异常神经模式知之甚少。该项目将利用一个创新的框架，整合多尺度记录和刺激，探索APD并阐明其潜在机制。该项目联合了一个多学科研究团队，包括神经信号处理、神经科学、精神病学和深度学习方面的专家。拟议的工作将在现实环境中开发计算、数据驱动的方法。这些研究将研究具有不同时空特性的多模态信号，并结合APD精神病的神经影像学研究。除了这项建议的科学影响外，拟议的工作将通过解决神经科学和精神病学方面的多个现有差距来促进国民健康。教育和推广计划将为女性和代表性不足的少数族裔提供培训机会，促进美国东北部STEM的多样性。这个项目有三个重点。所有提议的框架都是数据驱动的，并将在健康对照者和精神分裂症患者身上进行测试，其中APD是精神分裂症的核心特征。第一个重点将发展一种计算统计方法，通过在听觉任务相关设置中嵌套的多模态方法，量化血流动力学次慢振荡（使用fNIRS）和电高频振荡（使用EEG）之间的分层耦合。第二个重点介绍了一种创新的多模态数据融合方法，利用电和血管动力学的互补优势，对APD进行综合理解。这将能够识别APD背后的跨主题和主题内信号。第三个重点将超越功能调查，深入大规模网络的因果动态。该研究将建立融合的因果模型，以确定APD的特定主题因果模式，并创建个性化的空间目标映射，以获得最佳的部位刺激。异常因果模式的精确位置将成为经颅直流电刺激（tDCS）的目标。项目成果包括引入创新的计算数据融合方法来跨越不同的时空尺度；通过新的神经信息处理发现APD的潜在特征和因果模式；深入了解APD的层次机制，并了解APD的皮质调节特性。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Auditory processing dysfunction (APD) is a common feature of many types of psychosis, including schizophrenia, and is associated with multiple core symptoms, including auditory verbal hallucinations (i.e. hearing voices). Despite APD’s high prevalence, affecting up to 80% of the psychotic population, pharmaceutical therapy is ineffective: 70% of patients either have undesirable side effects or experience persistent symptoms despite treatment. A non-pharmacological treatment strategy, such as neuromodulation (targeted stimulation of nerves), would meet an important medical need. Although neuromodulation has recently emerged as a plausible therapeutic tool for a range of neuropsychological conditions, little is understood of the abnormal neural patterns underlying APD. This project will utilize an innovative framework, integrating multiscale recording and stimulation, to explore APD and to elucidate its underlying mechanisms. The project unites a multidisciplinary team of researchers, including experts in neural signal processing, neuroscience, psychiatry, and deep learning. The proposed work will develop computational, data-driven approaches in real-world settings. These will investigate multimodal signals with distinct spatiotemporal properties, integrated with a neuroimaging study of psychosis with APD. In addition to the scientific impacts of this proposal, the proposed work will advance national health by addressing multiple existing gaps in neuroscience and psychiatry. The educational and outreach plans will provide training opportunities for women and under-represented minoroties, promoting STEM diversity in the Northeastern United States.This project has three main thrusts. All of the proposed frameworks are data-driven and will be tested on healthy controls and patients with schizophrenia, in whom APD is a core feature. The first thrust will develop a computational statistical approach to quantify hierarchical couplings between hemodynamic infra-slow oscillations (using fNIRS), and electrical high-frequency oscillations (using EEG), through a nested multimodal approach in auditory task-related settings. The second thrust introduces an innovative multimodal data fusion approach to exploit complementary strengths from electrical and vascular dynamics, toward an integrative understanding of APD. This will enable identification of across-subject and within-subject signals underlying APD. The third thrust will extent beyond functional investigations and into causal dynamics across large-scale networks. The research will develop fused causal models, to identify subject-specific causal patterns of APD, and to create individualized spatial target mapping for optimal site stimulation. The precise locations of aberrant causal patterns will be targets for transcranial direct-current stimulation (tDCS). Project outcomes include the introduction of an innovative computational data fusion approach to bridge distinct spatiotemporal scales; discovery of latent signatures and causal patterns of APD through novel neural information processing; insight into APD hierarchical mechanisms, and understanding of the cortical modulatory properties of APD.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.