3D IVUS-EM Reconstruction and CT Registration Data for "learned ultrasound segmentation and deformable CT fusion for augmented reality endovascular surgery", Dillon et al.

This dataset contains the raw IVUS, EM, ECG intraoperative data and corresponding CT ground truth data for 7 in vitro phantoms and 3 in vivo animals - image_npys - ultrasound images in numpy format- image_jpgs - ultrasound images in jpg format- EM - 4 x 4 transforms, formatted as numpy files (T_W_EM)- calibration_parameters_ivus - 4 x 4 spatial transform, describing offset of the IVUS probe relative to the EM sensor for that experiment (T_EM_IVUS)- preop_data - contains STL mesh file and centerline data extracted from 3D Slicer which was registered to the 3D IVUS/EM reconstructions As described in our manuscript, 3D reconstructions can be obtained by using the equation X_3D = T_W_EM * T_EM_IVUS * X_2D, where X_2D are points segmented from the ultrasound image for the in vivo data, there is additional parameters for ECG gating:- ecgtimes - timestamp for each ecg reading measured in ROS- imagetimes - timestamp for each image measured in ROS- ecg_signal - ecg signal intensity at each timepoint (1 or 0); used for gating