Examining the electroencephalographic fingerprint of default mode network hyperconnectivity for scalable and personalized neurofeedback in schizophrenia

项目摘要 幻听（AH）是精神分裂症（SZ）的核心症状之一， 痛苦和残疾的根源。三分之一的SZ患者经历了药物耐药AH， 必须制定替代/补充治疗策略。研究人员开始 我理解精神疾病是如何与复杂的交流模式中的特定变化相关联的 由于磁共振成像（MRI）的新进展，不同的大脑区域之间的联系。特别是， 功能磁共振成像（fMRI）数据采集和计算分析的创新，使 现在可以以个性化的方式可靠地映射大脑网络的功能神经解剖学， 潜在的途径，以确定独特的和个性化的神经治疗目标。此外，现在有可能 量身定制个人和非侵入性干预，以帮助患者正常化内部和之间的沟通 使用实时神经反馈的复杂大脑网络-患者通过观察并学习调节 他们大脑活动的某些方面。AH的特征是内在功能连接性升高 在默认模式网络（DMN）内，以及DMN和其他大规模网络（如额顶叶）之间， 控制网络（FPCN）和听觉皮层（即，上级颞回（STG））。我们最近开发了一种 创新实时功能磁共振成像电路神经反馈（rt-fMRI-NF）范例，使人们观察视觉显示 持续的DMN激活水平，并使用正念作为一种策略来自愿调节这种差异。我们 研究表明rt-fMRI-NF降低了DMN超连接性并增加了DMN-FPCN超连接性关系， 在被诊断为SZ的成年人中，AH的相关减少。不幸的是，要瞄准主脑 神经网络在神经精神疾病中功能异常，神经反馈目前依赖于fMRI 这是一种昂贵的手术，涉及复杂的设置和患者负担。因为频率- 在头皮上记录的脑电图（EEG）信号的特定分量可以用作以下的相关物： fMRI活动模式，包括DMN活动和连接。在这里，我们建议验证EEG 使用同步EEG-fMRI研究涉及AH的DMN相互作用的相关性，并开发EEG 这些功能磁共振成像网络动态的“指纹”。因此，我们将扩展我们成功的rt-fMRI-NF策略， 创新性地增加了同步EEG测量。我们将应用最新的个性化技术 fMRI功能网络映射定义脑电信号特征，预测和优化脑电信号指纹 使用机器学习的生物信号，这可能会导致未来的个性化，网络- 脑电神经反馈回路治疗SZ型AH的临床研究这项研究将提供关键的技术创新， 导致新颖和可扩展临床应用。我们将对40名SZ和AH患者进行丰富（>30分钟）采样 与同步EEG-fMRI开发DMN动态的开创性和个性化EEG指纹， 因此能够为患者提供可扩展形式精确的基于网络的神经反馈训练。

PROJECT ABSTRACT Auditory hallucinations (AHs) are one of the core symptoms of schizophrenia (SZ) and constitute a significant source of suffering and disability. One third of SZ patients experience pharmacology-resistant AHs, such that it is imperative to develop alternative/complementary treatment strategies. Researchers are beginning to appreciate how mental illnesses are associated with specific changes in the complex patterns of communication between different brain regions thanks to new advances in Magnetic Resonance Imaging (MRI). In particular, innovations in functional Magnetic Resonance Imaging (fMRI) data acquisition and computational analysis, make it now possible to reliably map the functional neuroanatomy of brain networks in a personalized way, offering a potential avenue for identifying unique and individualized neurotherapeutic targets. Moreover, it is now possible to tailor a personal and noninvasive intervention to help patients normalize communication within and between complex brain networks using real-time neurofeedback— whereby patients observe and learn to regulate selected aspects of their own brain activity—. AHs are characterized by elevated intrinsic functional connectivity within the default mode network (DMN) and between DMN and other large-scale networks like the frontoparietal control network (FPCN) and auditory cortices (i.e., superior temporal gyrus (STG)). We recently developed an innovative real-time fMRI circuit neurofeedback (rt-fMRI-NF) paradigm whereby people observe a visual display of ongoing DMN activation levels and use mindfulness as a strategy to volitionally regulate this difference. Our research has shown that rt-fMRI-NF reduces DMN hyperconnectivity and increases DMN-FPCN anticorrelations, with a correlated reduction of AHs among adults diagnosed with SZ. Unfortunately, to target the major brain networks that function abnormally in neuropsychiatric conditions, neurofeedback currently relies on fMRI technology, which is an expensive procedure involving a complex setup and patient burden. Since frequency- specific components of electroencephalography (EEG) signals recorded on the scalp can serve as correlates of fMRI activity patterns, including DMN activity and connectivity. Here we propose to validate the EEG correlates of DMN interactions implicated in AHs using concurrent EEG-fMRI and to develop an EEG “fingerprint” of these fMRI network dynamics. Hence, we will expand our successful rt-fMRI-NF strategy with the innovative addition of concurrent EEG measurements. We will apply the latest advances in personalized fMRI functional network mapping to define the features of EEG signal to predict and optimize the EEG fingerprint of fMRI activity using advances in machine learning for bio-signals that may lead to future personalized, network- based EEG neurofeedback circuit therapy for AHs in SZ. This study will offer key technical innovations that could lead to novel and scalable clinical applications. We will richly (>30 minutes) sample 40 patients with SZ and AHs with simultaneous EEG-fMRI to develop a pioneering and personalized EEG fingerprint of DMN dynamics and so enable a scalable form of accurate network-based neurofeedback training to patients.