The Mouse Grimace Scale (MGS) [1] has emerged as a valuable tool for pain assessment in mice by analyzing facial features. This study aims to automate and enhance this process through the development of an innovative image capture system, utilizing near-infrared (NIR) spectrum illumination for less intrusive and clearer facial images. Real-time pain assessment is achieved by precise segmentation of the eye, ear, and muzzle areas using fine-tuned Mask R-CNN. Furthermore, a comprehensive dataset of over 3000 high-quality images, created by us, was employed for training and validation, resulting in improved accuracy and reliability of our pain evaluation system. These contributions not only advance the field of pain evaluation in mice but also address a limitation in a previous study that solely focused on assessing pain based on the eye area [2]. By incorporating additional facial regions, our approach provides a more comprehensive understanding of pain manifestations in mice, enabling a more nuanced and accurate assessment. These findings open up new avenues for further research and development in pain evaluation methodologies.
[1]
A. Schwing,et al.
Masked-attention Mask Transformer for Universal Image Segmentation
,
2021,
2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).
[2]
George Papandreou,et al.
Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation
,
2018,
ECCV.
[3]
Ross B. Girshick,et al.
Mask R-CNN
,
2017,
1703.06870.
[4]
Thomas Brox,et al.
U-Net: Convolutional Networks for Biomedical Image Segmentation
,
2015,
MICCAI.
[5]
K. Craig,et al.
Coding of facial expressions of pain in the laboratory mouse
,
2010,
Nature Methods.
[6]
Justus Schock,et al.
Introducing CNN-Based Mouse Grim Scale Analysis for Fully Automated Image-Based Assessment of Distress in Laboratory Mice
,
2018,
VCBM@MICCAI.