A neuropsychological approach to dissect face perception and perceptual expertise: Behavioural data

This collection contains the response time and accuracy data during the training sessions and during the pre-training and post-training test sessions. We used an inversion task for the test sessions. During all sessions, subjects were responding to faces or novel objects.
The data for this study are organised into four collections. The first, 'Visual stimuli', contains the visual stimuli used throughout the study. These includes faces and novel three-dimensional (3D) objects rendered from different viewpoints. This collection also contains the 3D model and Matlab scripts to help create more stimuli. The second, 'EEG data', contains the processed EEG responses to faces and novel objects during the pre-training and post-training test sessions. We used the fast periodic visual stimulation (FPVS) paradigm. Subjects performed a fixation task during EEG data collection. The last collection, "'FMRI data', contains the raw functional data for brain responses to faces and novel objects. In this phase, we adapted the FPVS paradigm used in the EEG study to the FMRI study. Subjects performed a fixation task during FMRI data collection. We also acquired structural and diffusion imaging data. (Find the other collections under 'Related Resources')<p>Recognising faces is at the heart of human social interactions. By adulthood, people are very good at extracting identity, sex, race, emotions, and social signals from faces. Therefore, impairments to this ability can drastically reduce their quality of life.

The aim of this project is to investigate the neural mechanisms underlying people’s ability to process faces and how these mechanisms adapt with experience. The approach is to test whether individuals with prosopagnosia can acquire expertise of novel non-face objects through training. These individuals had head trauma during adulthood that lead to damage in specific brain regions. These regions are thought to process only faces and no other object categories. However, these regions may be more generally involved in processing object categories for which people have expertise (eg, bird experts). In addition to neurological case studies, volunteers will also go through the training. Their brain will be scanned using magnetic resonance imaging to determine how the putative face-specific regions change over the course of training. Overall, the results will have an impact on clinical populations which can result in face recognition deficits, such as Alzheimer’s disease, stroke patients, and developmental disorders that affect social interactions (eg, Autism).</p>

本数据集收录了训练阶段，以及预训练、后训练测试阶段中的响应时长与准确率数据。测试阶段采用反转任务范式。所有实验阶段中，被试均需对人脸或新型物体做出反应。
本研究的数据分为四个数据集板块：第一板块为「视觉刺激集（Visual stimuli）」，收录了本研究全程使用的视觉刺激材料，包括从不同视角渲染的人脸与新型三维（3D）物体，同时还包含用于生成更多刺激材料的三维模型与MATLAB脚本。第二板块为「脑电图（EEG）数据」，收录了预训练与后训练测试阶段中，被试对人脸及新型物体产生的经过预处理的脑电图响应。本研究采用快速周期性视觉刺激（fast periodic visual stimulation, FPVS）范式，在脑电图数据采集过程中，被试需完成注视任务。第四板块为「功能磁共振成像（fMRI）数据」，收录了被试对人脸及新型物体产生脑响应的原始功能成像数据。本阶段将脑电图研究中使用的FPVS范式适配至功能磁共振成像研究中，数据采集过程中被试仍需完成注视任务。此外，本研究还采集了结构成像与弥散张量成像数据。其余未详细说明的数据集板块，请参见「相关资源（Related Resources）」。
人脸识别是人类社会互动的核心能力之一。至成年阶段，人类已能够熟练从人脸中提取身份、性别、种族、情绪及社会信号。因此，该能力受损会严重降低个体的生活质量。
本项目旨在探究人类加工人脸的神经机制，以及这些机制如何随训练经验发生适应性改变。具体研究路径为验证面孔失认症（prosopagnosia）患者能否通过训练获得新型非人脸物体的专业识别能力。此类患者在成年时期因头部创伤导致特定脑区受损，该类脑区曾被认为仅负责加工人脸而非其他物体类别，但后续研究表明，其或许可更广泛地参与加工人类具备专业知识的物体类别（例如鸟类识别专家的相关脑区）。除神经病例研究外，本研究还将招募健康志愿者参与训练，并通过磁共振成像扫描其脑部活动，以明确假定的人脸特异性脑区在训练过程中的动态变化。总体而言，本研究结果将对存在人脸识别障碍的临床人群具有重要意义，包括阿尔茨海默病（Alzheimer’s disease）患者、脑卒中患者，以及影响社会互动的发育障碍群体（例如孤独症（Autism）患者）。