Welcome to CSGO’s documentation!

Note

This is the source code documentation page. Looking for a quick start guide? Please visit https://github.com/QBRC/CSGO.

Introduction

Cell Segmentation with Globally Optimized boundaries (CSGO) is a whole-cell segmentation pipeline tailored specifically to the H&E domain. There are three main classes used in this pipeline - CSGO, HD-Yolo, and UNet. The detailed function documentation is listed below.

Contents

• CSGO
  • CSGO
    • CSGO.convert_resolution_to_mpp()
    • CSGO.find_missed_nuclei()
    • CSGO.membrane_detection()
    • CSGO.predict_membrane_from_patch()
    • CSGO.segment()
    • CSGO.to_device()
    • CSGO.unet_init()
    • CSGO.watershed()
• HD-Yolo
  • yoloStandalone
    • yoloStandalone.args_init()
    • yoloStandalone.run_inference()
• UNet
  • Conv2d
  • Conv2dBNReLU
  • ConvBNReLU
  • IoU
    • IoU.forward()
  • ResizeConv2d
  • SoftDiceLoss
    • SoftDiceLoss.forward()
  • SubPixelConv2d
    • SubPixelConv2d.icnr_()
  • UNet
    • UNet.forward()
    • UNet.get_encoder()
  • UNetFeatures
    • UNetFeatures.forward()
  • get_norm_layer_and_bias()