THINGS-fMRI

# THINGS-fMRI

Understanding object representations visual and semantic processing of objects requires a broad, comprehensive sampling of the objects in our visual world
with dense measurements of brain activity and behavior. This densely sampled fMRI dataset is part of THINGS-data, a multimodal collection
of large-scale datasets comprising functional MRI, magnetoencephalographic recordings, and 4.70 million behavioral judgments in response to thousands of photographic images
for up to 1,854 object concepts. THINGS-data is unique in its breadth of richly-annotated objects, allowing for testing countless novel hypotheses at scale while assessing
the reproducibility of previous findings. The multimodal data allows for studying both the temporal and spatial dynamics of object representations and their relationship
to behavior and additionally provides the means for combining these datasets for novel insights into object processing. THINGS-data constitutes the core release of
the [THINGS initiative](https://things-initiative.org) for bridging the gap between disciplines and the advancement of cognitive neuroscience.

# Dataset overview
We collected extensively sampled object representations using functional MRI (fMRI). To this end, we drew on the THINGS database (Hebart et al., 2019), a richly-annotated database of 1,854 object concepts representative of the American English language which contains 26,107 manually-curated naturalistic object images.

During the fMRI experiment, participants were shown a representative subset of THINGS images, spread across 12 separate sessions (N=3, 8740 unique images of 720 objects). Images were shown in fast succession (4.5s), and participants were instructed to maintain central fixation. To ensure engagement, participants performed an oddball detection task responding to occasional artificially-generated images. A subset of images (n=100) were shown repeatedly in each session.

Beyond the core functional imaging data in response to THINGS images, additional structural and functional imaging data were gathered. We collected high-resolution anatomical images (T1- and T2-weighted), measures of brain vasculature (Time-of-Flight angiography, T2*-weighted) and gradient-echo field maps. In addition, we ran a functional localizer to identify numerous functionally specific brain regions, a retinotopic localizer for estimating population receptive fields, and an additional run without external stimulation for estimating resting-state functional connectivity.

Besides raw data this datasets holds

- brainmasks (fmriprep)
- cortical flat maps (pycoretx_filestore)
- single trial response estimations (ICA-betas)

More derivatives can be found on [figshare](https://doi.org/10.25452/figshare.plus.c.6161151.v1).

# Provenance

Provenance information is given in 'dataset_description.json' as well as in the [paper](https://doi.org/10.7554/eLife.82580) and preprocessing and analysis code is shared on [GitHub](https://github.com/ViCCo-Group/THINGS-data).

# THINGS-fMRI
要理解物体表征、视觉与语义加工，需对视觉世界中的物体进行广泛且全面的采样，并辅以脑活动与行为的高密度测量。本高密度采样的功能磁共振成像（functional MRI, fMRI）数据集隶属于THINGS-data——一个多模态大规模数据集集合，涵盖功能磁共振成像、脑磁图（magnetoencephalographic）记录，以及针对多达1854个物体概念的数千张摄影图像所收集的470万份行为判断数据。THINGS-data的独特之处在于其拥有大量经过精细标注的物体，可在大规模层面验证无数全新假说，同时也能对既往研究结果的可重复性进行评估。该多模态数据集可用于研究物体表征的时空动态及其与行为的关联，同时还为整合这些数据集以获得物体加工的全新见解提供了可能。THINGS-data是[THINGS研究计划](https://things-initiative.org)的核心发布成果，旨在打破学科壁垒，推动认知神经科学的发展。

# 数据集概览
本研究通过功能磁共振成像（fMRI）采集了经过高密度采样的物体表征数据。为此，我们采用了THINGS数据库（Hebart等人，2019）——该数据库包含26107张经人工精心筛选的自然物体图像，涵盖1854个符合美式英语语境的典型物体概念，且所有物体均经过精细标注。

功能磁共振成像实验中，我们向被试呈现THINGS数据库中具有代表性的图像子集，实验共分为12个独立扫描段，涉及3名被试、720个物体的8740张独特图像。图像以快速序列方式呈现（每张4.5秒），要求被试始终保持中央注视。为确保被试保持任务专注度，实验设置了异常刺激检测任务：被试需对偶尔出现的人工生成图像做出反应。每一个扫描段中都会重复呈现100张图像子集。

除针对THINGS图像的核心功能成像数据外，本数据集还包含额外的结构与功能成像数据。我们采集了高分辨率解剖图像（T1加权与T2加权成像）、脑血管测量数据（飞行时间血管造影、T2*加权成像）以及梯度回波场图。此外，实验还开展了多项定位扫描：包括识别多个功能特异性脑区的功能定位任务、用于估算群体感受野的视网膜定位任务，以及一段无外部刺激的扫描以获取静息态功能连接数据。

除原始数据外，本数据集还包含：
- 脑掩膜（fmriprep）
- 皮层平铺图（pycoretx_filestore）
- 单试次反应估计值（ICA-betas）

更多衍生数据可在[figshare](https://doi.org/10.25452/figshare.plus.c.6161151.v1)获取。

# 数据溯源
数据溯源信息可在'dataset_description.json'文件以及[相关论文](https://doi.org/10.7554/eLife.82580)中获取，预处理与分析代码已共享至[GitHub](https://github.com/ViCCo-Group/THINGS-data)

搜集汇总

背景与挑战

背景概述

THINGS-fMRI数据集是THINGS-data多模态数据集的核心组成部分，专注于功能磁共振成像（fMRI）数据，旨在研究物体表征的视觉和语义处理。该数据集包含对1,854个物体概念的广泛采样，涉及3名参与者的12个会话，收集了密集的fMRI响应数据以及额外的结构性和功能性成像数据，如解剖图像和功能定位器。其特点在于多模态集成和丰富衍生数据（如脑掩膜和单试验响应估计），支持大规模、可重复的神经科学研究。