Inventions, Vol. 9, page 25: Optimizing zebrafish skin ablation laser methods and developing deep learning-based skin wound size measurement methods
Inventionsdoi:10.3390/inventions9020025
Author: Petrus Siregar Yi-Shan Liu Franelyne P. Casuga Ching-Yu Huang Kelvin H.-C. Click to download and save Chen Rong – Qin Huangzhi – Xin Hongyi – Kailin Zhongde Xiaohong – Yulin mp3 youtube com
The skin plays an important role as a defense mechanism for organisms such as humans or animals against environmental pathogens. Once the integrity of the skin is compromised by a wound, pathogens can easily penetrate deeper into the body, inducing disease. In this way, it is important that the skin regenerates quickly after injury to restore its protective barrier function. Traditionally, scientists have used rodents or mammals as experimental animals to study skin wound healing. However, due to animal welfare concerns and the increasing cost of laboratory animals such as rodents, scientists have considered alternative methods of implementing replacement, reduction, and improvement (3R) in experiments. Moreover, several previous studies on fish skin wound healing have used relatively expensive medical-grade lasers, and the calculation of wound area is inefficient, leading to human judgment errors. Therefore, this study aimed to develop a new alternative model of skin wound healing using zebrafish and a new rapid and effective method as an alternative method to study skin wound healing. First, to implement the 3R concept, pain in test zebrafish was assessed by using 3D motion assay. Subsequently, the behavioral data obtained were analyzed using Kruskal-Wallis test and Dunn’s multiple comparison test; 3 watts was later selected as the laser power because the wounds caused by the laser at this power did not significantly change the swimming behavior of zebrafish. In addition, we also used a laser engraving machine to optimize the experimental conditions for zebrafish skin wound healing, which can produce skin wounds with high reproducibility in size and depth. Zebrafish were then tested for wound closure using a two-way ANOVA analysis and expressed as wound closure percentages of 25%, 50%, and 75%. After leaving a wound on the skin of the zebrafish, images of the wound are collected and trained with deep learning through a convolutional neural network (CNN) (Mask-RCNN or U-Net) so that the computer can calculate the area of the skin to wound the wound in an automated manner. Using ImageJ manual counting as the gold standard, we found that U-Net performed better than Mask RCNN in zebrafish skin wound judgment. For proof-of-concept, U-Net trained models were applied to study and determine the effects of different temperatures and antioxidant administration on skin wound healing kinetics. The results showed a significant positive correlation between wound closure speed and exposure to different temperatures and taking antioxidants. Taken together, the laser-based skin ablation and deep learning-based wound size measurement methods reported in this study provide for the first time a faster, reliable, and less painful protocol for skin wound healing in zebrafish.