Fetoscopy Placenta Dataset

<p dir="ltr"><b>First Fetoscope Placenta Dataset for In-Vivo Vessel Segmentation and Mosaicking</b></p><h3><b>Overview</b></h3><p dir="ltr">The fetoscope placenta dataset described in [<a href="https://link.springer.com/chapter/10.1007/978-3-030-59716-0_73" rel="noreferrer" target="_blank">Bano et al. MICCAI2020</a>] consists of in-vivo fetoscopic imagery from twin-to-twin transfusion syndrome (TTTS) procedures. It includes:</p><ul><li><b>483 frames</b> from <b>6 fetoscopic videos</b> with expert <b>ground-truth placental vessel segmentation</b> annotations.</li><li><b>6 additional continuous fetoscopic video clips</b> (total <b>950 frames</b>) without manual labels but accompanied by <b>predicted vessel segmentation maps</b> generated via leave-one-out cross-validation.</li></ul><p dir="ltr">The dataset is intended for research on <b>placental vessel segmentation</b> and <b>fetoscopic video mosaicking/registration </b>under realistic surgical conditions (limited field of view, specularities, and variable visibility).</p><h3><b>Citing the Dataset</b></h3><p dir="ltr">Cite [<a href="https://link.springer.com/chapter/10.1007/978-3-030-59716-0_73" rel="noreferrer" target="_blank">Bano et al. MICCAI2020</a>] whenever research making use of this dataset is reported in any academic publication or research report.</p>