tkhangg0910/brain_data_split_v2

该数据集是一个医学影像数据集，包含患者ID、幻灯片信息、交付物、医生信息、边界框坐标（标准化）、多语言问答对（Q1-Q4和A1-A4）、图像数据、多种语言的翻译版本、不同类型的掩膜（GCA_mask、Koedam_mask、MTA_mask）以及标签列表等。数据集分为训练集和测试集，训练集有1031个样本，测试集有511个样本。数据集中还包含了不同类别的分布情况，如状态分布（Moderate-Dementia、Non-Dementia、Mild-Dementia）、GCA分布、Koedam分布、MTA分布等。此外，还提供了边界框统计信息，如平均每个样本的边界框数量、最大边界框数量等。

This dataset is a medical imaging dataset containing patient IDs, slide information, deliverables, doctor information, normalized bounding box coordinates, multilingual Q&A pairs (Q1-Q4 and A1-A4), image data, translations in multiple languages, different types of masks (GCA_mask, Koedam_mask, MTA_mask), and label lists. The dataset is divided into training and test sets, with 1031 samples in the training set and 511 samples in the test set. The dataset also includes distribution information for different categories, such as status distribution (Moderate-Dementia, Non-Dementia, Mild-Dementia), GCA distribution, Koedam distribution, MTA distribution, etc. Additionally, it provides bounding box statistics, such as the average number of bounding boxes per sample and the maximum number of bounding boxes.