Table 2 Comparative evaluation of microglial detection and segmentation methods

From: StainAI: quantitative mapping of stained microglia and insights into brain-wide neuroinflammation and therapeutic effects in cardiac arrest

 

Cell Detection

Cell Segmentation

Model

Control

Injured

Control

Injured

Mask R-CNN

0.695

0.686

0.336

0.541

YOLACT

0.663

0.640

0.309

0.508

YOLO+UNet

0.762

0.715

0.763

0.712

  1. The performance of Mask R-CNN, YOLACT, and YOLO+UNet was evaluated by their mean average precision at a 50% intersection over union (mAP50) using a dataset comprising 138 microglial images annotated from both control (n = 72) and injured (n = 66) brains.
Back to article page