Dataset Hub

The dataset contains small‑size pathology images with corresponding labels indicating the presence of tumor tissue. Images are 96 × 96 pixels and were provided as part of a Kaggle competition.

The dataset contains images and labels with three categories: Trans‑cerebellum, Trans‑thalamic, and Trans‑ventricular. It includes only a training split with 2,949 samples, total size 602,951,851.225 bytes, download size 571,013,578 bytes.

This dataset originates from Guangdong Provincial People's Hospital and contains cardiac CT images from 92 patients. Image dimensions are 512 × 512 × (137‑251) with a typical voxel size of 0.25 × 0.25 × 0.5 mm³. Experienced radiologists performed precise annotations, taking approximately one hour per CT scan. The dataset includes ten labels: left ventricle, right ventricle, left atrium, right atrium, aorta, pulmonary artery, myocardium, superior vena cava, inferior vena cava, and pulmonary veins.

The BCCD dataset is a small‑scale collection for blood cell detection, containing three label types: red blood cells (RBC), white blood cells (WBC), and platelets. Images are in JPEG format with a resolution of 640 × 480 pixels, and annotations are provided in VOC‑style XML files.

We contributed to the development of the VQA‑RAD dataset by acquiring radiology reports. Our work involved collecting and validating these reports to ensure clear structure and accurate textual information corresponding to each image.

The Ultralytics Brain‑tumor Dataset contains medical images from MRI or CT scans, focusing on brain tumor detection. It provides information on tumor presence, location, and characteristics, which is essential for training computer‑vision algorithms to automate brain tumor identification, aiding early diagnosis and treatment planning.