StarDist_BF_Neutrophil_dataset

<p>This repository includes a StarDist deep learning model and its training and validation datasets for detecting neutrophils perfused over an endothelial cell monolayer. The model was trained on 36 manually annotated images, achieving an average F1 Score of 0.969. The dataset and model are intended for use in biomedical research, particularly for analyzing interactions between neutrophils and endothelial cells.</p> <h3>Specifications</h3> <ul> <li> <p>Model: StarDist for neutrophil detection on endothelial cells</p> </li> <li> <p>Training Dataset:</p> </li> <ul> <li> <p>Number of Images: 36 paired brightfield microscopy images and label masks</p> </li> <li> <p>Microscope: Nikon Eclipse Ti2-E, 20x objective</p> </li> <li> <p>Data Type: Brightfield microscopy images with manually segmented masks</p> </li> <li> <p>File Format: TIFF (.tif)</p> </li> <ul> <li> <p>Brightfield Images: 16-bit</p> </li> <li> <p>Masks: 8-bit</p> </li> </ul> <li> <p>Image Size: 1024 x 1022 pixels (Pixel size: 650 nm)</p> </li> </ul> <li> <p>Training Parameters:</p> </li> <ul> <li> <p>Epochs: 400</p> </li> <li> <p>Patch Size: 992 x 992 pixels</p> </li> <li> <p>Batch Size: 2</p> </li> </ul> <li> <p>Performance:</p> </li> <ul> <li> <p>Average F1 Score: 0.969</p> </li> <li> <p>Average IoU: 0.914</p> </li> </ul> <li> <p>Model Training: Conducted using ZeroCostDL4Mic (<a href="https://github.com/HenriquesLab/ZeroCostDL4Mic/wiki/Stardist">https://github.com/HenriquesLab/ZeroCostDL4Mic/wiki</a>)</p> </li> </ul> <div> <h3>Reference</h3> <div><strong>Fast label-free live imaging reveals key roles of flow dynamics and CD44-HA interaction in cancer cell arrest on endothelial monolayers</strong></div> </div> <div>Gautier&nbsp;Follain,&nbsp;Sujan&nbsp;Ghimire,&nbsp;Joanna W.&nbsp;Pylv&auml;n&auml;inen,&nbsp;Monika&nbsp;Vaitkevičiūtė,&nbsp;Diana&nbsp;Wurzinger,&nbsp;Camilo&nbsp;Guzm&aacute;n,&nbsp;James RW&nbsp;Conway,&nbsp;Michal&nbsp;Dibus,&nbsp;Sanna&nbsp;Oikari,&nbsp;Kirsi&nbsp;Rilla,&nbsp;Marko&nbsp;Salmi,&nbsp;Johanna&nbsp;Ivaska,&nbsp;Guillaume&nbsp;Jacquemet</div> <div>bioRxiv&nbsp;2024.09.30.615654;&nbsp;doi:&nbsp;<a href="https://www.biorxiv.org/content/10.1101/2024.09.30.615654v1">https://doi.org/10.1101/2024.09.30.615654</a></div>