About Dataset

This dataset is for training a multiple object detector in X-ray fluoroscopic images. Images were acquired in 72 different clinical cases. There were 6,533 images from 9 transcatheter aortic valve replacement (TAVR) procedures, 250 images from one atrial fibrillation (AF) ablation procedure guided by X-ray and transesophageal echo (TOE) images and 5,655 images from 62 standard ablation procedures.

To test the data labelling, please use LoadBoxTest.py. It will draw rotated bounding boxes on the target objects.

The codes are available at https://github.com/YingLiangMa/AttWire