AIM-CICs: an automatic identification method for cell-in-cell structures based on convolutional neural network  

作　　者：Meng Tang Yan Su Wei Zhao Zubiao Niu Banzhan Ruan Qinqin Li You Zheng Chenxi Wang Bo Zhang Fuxiang Zhou Xiaoning Wang Hongyan Huang Hanping Shi Qiang Sun 

机构地区：[1]Beijing Shijitan Hospital of Capital Medical University,Beijing 100038,China [2]Laboratory of Cell Engineering,Institute of Biotechnology,Research Unit of Cell Death Mechanism,Chinese Academy of Medical Science,2021RU008,Beijing 100071,China [3]School of Mathematical Sciences,Peking University,Beijing 100871,China [4]Department of Radiation and Medical Oncology,Hubei Key Laboratory of Tumor Biological Behaviors,Hubei Clinical Cancer Study Center,Zhongnan Hospital,Wuhan University,Wuhan 430071,China [5]National Clinic Center of Geriatric&State Key Laboratory of Kidney,Chinese PLA General Hospital,Beijing 100853,China [6]Comprehensive Oncology Department,National Cancer Center/Cancer Hospital,Chinese Academy of Medical Sciences and Peking Union Medical College,Beijing 100021,China

出　　处：《Journal of Molecular Cell Biology》2022年第6期57-67,共11页分子细胞生物学报（英文版）

基　　金：This workwas supported by Beijing Municipal Natural Science Foundation(KZ202110025029 to H.H.);the National Key R&D Program of China(2022YFC3600100 to Q.S.and H.H.);the National Natural Science Foundation of China(32100608 to C.W.,82002918 and 31970685 to Q.S.);Beijing Municipal Administration of Hospitals Incubating Program(PX2021033 to H.H.);Beijing Postdoctoral Research Foundation(2021-ZZ-027 to M.T.).

摘　　要：Whereas biochemical markers are available for most types of cell death, current studies on non-autonomous cell death by entosis rely strictly on the identification of cell-in-cell structures (CICs), a unique morphological readout that can only be quantified manually at present. Moreover, the manual CIC quantification is generally over-simplified as CIC counts, which represents a major hurdle against profound mechanistic investigations. In this study, we take advantage of artificial intelligence technology to develop an automatic identification method for CICs (AIM-CICs), which performs comprehensive CIC analysis in an automated and efficient way. The AIM-CICs, developed on the algorithm of convolutional neural network, can not only differentiate between CICs and non-CICs (the area under the receiver operating characteristic curve (AUC) > 0.99), but also accurately categorize CICs into five subclasses based on CIC stages and cell number involved (AUC > 0.97 for all subclasses). The application of AIM-CICs would systemically fuel research on CIC-mediated cell death, such as high-throughput screening.

关 键 词：cell-in-cell structure artificial intelligence AIM-CICs cell death entosis convolutional neural network 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]