OctoMed/BreastMNIST

该数据集包含用于二元诊断分类的乳腺超声图像。任务涉及将每张超声图像分类为恶性或正常/良性。本任务中每个样本收集了16条由GPT-4o生成的理由追踪，存储在`responses`列中。数据集基于https://medmnist.com的原始数据源构建。数据字段包括：`question`（关于乳腺超声诊断的分类问题）、`options`（代表诊断类别的多项选择选项）、`answer`（正确诊断）、`image`（乳腺超声图像）和`responses`（训练集中的模型推理响应）。数据集分为训练集（含模型响应）和测试集（用于评估）。

This dataset contains breast ultrasound images for binary diagnostic classification. The task involves classifying each ultrasound image as malignant or normal/benign. 16 reasoning traces were collected for each example in this task by sampling with GPT-4o, available in the `responses` column. We greatly appreciate and build from the original data source available at https://medmnist.com. Data fields include: `question` (the classification question about breast ultrasound diagnosis), `options` (multiple choice options representing diagnostic categories), `answer` (the correct diagnosis), `image` (breast ultrasound image), and `responses` (model reasoning responses in train split). The dataset is split into training data (with model responses) and test data (for evaluation).