Automatedcell counting using deep learning in 1064-nm laser-induced cell apoptosis in pig cutaneous
作者单位：军事医学研究院
会议名称：《第十七届中国体视学与图像分析学术会议》
会议日期：2022年
学科分类：12[管理学] 1201[管理学-管理科学与工程(可授管理学、工学学位)] 1002[医学-临床医学] 08[工学] 081104[工学-模式识别与智能系统] 080203[工学-机械设计及理论] 0802[工学-机械工程] 0835[工学-软件工程] 0811[工学-控制科学与工程] 0812[工学-计算机科学与技术（可授工学、理学学位）] 100206[医学-皮肤病与性病学] 10[医学]
摘 要：Laser-induced cutaneous injury is primarily caused by the thermal ***-term radiation causes tissue degeneration,cell necrosis,and tissue coagulation or coagulation-type necrosis,tissue vaporization,or even carbonization and *** injury histological sectioning image is a difficult modality for automated cell segmentation and counting due to high noise and low *** such segmentation algorithms for cell counting and tracking typically yield more consistent results in other tissue histological sectioning such as tumor diagnosis due to better cell *** methods have shown that U-net based models can achieve state-of-the-art count and segmentation performance of tumor tissue slices,although all available methods continue to overly segment the collagen features and have difficulty capturing the entirety the *** propose a method for cell detection that requires annotated training small amount *** investigate the use of TransUNet[1],which merits both Transformers and U-Net,as a strong alternative for medical image segmentation,and show that it is able to improve the accuracy of Laser injury histological sectioning images in comparison to a baseline U-net *** work aims to obtain reliable cell segmentation and automated cell counting from specular microscopy images of both healthy and pathological *** investigate the biological effect of 1064-nm laser radiation,a comprehensive evaluation of histological sections were quantitatively *** radiation with different doses was performed on the back skin of *** trained a TransUNet model by extracting 460×500 pixel patches from 1064-nm laser-induced cell apoptosis in pig cutaneous images and the corresponding manual segmentation by a *** results show that the proposed method can deliver reliable feature segmentation enabling more accurate cell density estimations for assessing the apoptosis state of the pig cutaneous.