Neural mechanisms of sensory predictions in schizophrenia with hallucinations

描述（由申请人提供）：精神分裂症中的活动性精神病是世界上最严重和最繁重的医学疾病之一。然而，这种疾病中精神病症状的机制，如幻觉（即，在没有外部感觉刺激的情况下的异常感知），仍然难以捉摸。该K23申请提出了一项研究和培训计划，该计划将支持申请人成为NIH资助的独立研究者，专注于功能性神经影像学在精神分裂症精神病症状研究中的应用。本申请中的活动建立在候选人之前的培训基础上，并设置在一个资源丰富的环境中，这将促进他在以下方面的专业知识的发展：（1）先进的分析方法和功能磁共振成像（fMRI）研究的研究行为;（2）计算神经科学;（3）感知和认知研究;（4）精神分裂症的病理生理学和临床评估;（5）精神分裂症的诊断和治疗。以及（5）在对弱势群体进行科学研究时采取负责任和合乎道德的行为。结合功能性磁共振成像（fMRI）和计算建模，目前的研究计划旨在（1）定义精神分裂症中产生幻觉的神经机制;（2）基于预测编码，告知幻觉计算模型的发展，这是一个经过验证的理论框架，支持感觉系统在学习和预测外部环境中的幻觉中的作用。最重要的假设是，基于异常预测的感觉皮层功能衰减会导致感觉皮层过度活动，从而产生幻觉。为了验证这一假设，本研究将采用一种新的言语辨别功能磁共振成像范例，两组精神分裂症患者，那些活跃的，频繁的听觉言语幻觉和那些没有显着的幻觉史，和第三组健康对照。这种设计将允许测试感觉预测编码机制的功能障碍与精神分裂症患者幻觉的在线体验之间的直接联系，从而为这种疾病的精神病症状的神经生物学基础提供信息。总之，这个培训和研究计划将促进候选人的过渡到一个独立的研究生涯，并将有助于确定新的治疗目标难治性精神病。 相关性：预测编码框架和基于模型的fMRI在精神病性症状研究中的新应用将为精神病性症状的产生机制提供新的线索，从而填补精神分裂症研究中的一个重要空白。该项目将用于开发一种解释性的幻觉模型，可用于为人类和非人类动物模型的未来神经科学研究产生特定的，可测试的假设，并发现可能通过学习或药物治疗来改变的新目标（感觉预测缺陷）。

DESCRIPTION (provided by applicant): Active psychosis in schizophrenia is among the most severe and burdensome medical conditions worldwide. However, the mechanisms of psychotic symptoms in this disorder, such as hallucinations (i.e., abnormal percepts in the absence of external sensory stimuli), remain elusive. This K23 application presents a research and training program that will support the applicant on a path towards becoming an NIH-funded independent investigator focused on the application of functional neuroimaging to the study of psychotic symptoms in schizophrenia. The activities in this application build on the candidate's prior training and are set in a resource-rich environment that will foster his development of expertise in (1) advanced analytic methods and study conduct for functional magnetic resonance imaging (fMRI) research; (2) computational neuroscience; (3) perception and cognition research; (4) pathophysiology and clinical assessment of schizophrenia; and (5) responsible and ethical conduct in scientific research with vulnerable populations. Combining functional magnetic resonance imaging (fMRI) and computational modeling, the current research proposal seeks to (1) define the neural mechanisms that generate hallucinations in schizophrenia; and (2) inform the development of a computational model of hallucinations based on predictive coding, an empirically-validated theoretical framework that supports a role of sensory systems in learning and predicting regularities in the external environment. The overarching hypothesis is that abnormal prediction-based attenuation of sensory cortical function produces excessive activity in the sensory cortex that generates hallucinations. To test this hypothesis, the present study will employ a novel speech discrimination fMRI paradigm, two groups of patients with schizophrenia, those with active, frequent auditory verbal hallucinations and those without a significant history of hallucinations, and a third group of healthy controls. This design will allo for testing a direct link between dysfunction in sensory predictive-coding mechanisms and the online experience of hallucinations in patients with schizophrenia, and will thus inform the neurobiological basis of psychotic symptoms in this disorder. Together, this training and research program will facilitate the candidate's transition to an independent research career and will help identify new therapeutic targets for refractory psychosis. RELEVANCE: The novel application of the predictive-coding framework and model-based fMRI to the study of psychotic symptoms will shed new light on the mechanisms of generation of psychotic symptoms, thus filling an important gap in schizophrenia research. This project will serve to develop an explanatory model of hallucinations that can be used to generate specific, testable hypotheses for future neuroscience research in both humans and non-human animal models, and to uncover novel targets (sensory prediction deficits) likely modifiable by treatment via learning or pharmacotherapy.