Automated detection of apoptotic versus nonapoptotic cell death using label-free computational microscopy
Automated detection of apoptotic versus nonapoptotic cell death using label-free computational microscopy
复制标题
DOI：
10.1002/jbio.202100310
复制
发表时间：
2022-01-09
影响因子：
2.8
通讯作者：
Ray, Aniruddha
中科院分区：
文献类型：
--
作者：
Kabir, Md Alamgir;Kharel, Ashish;Ray, Aniruddha
细胞死亡机制的识别，特别是对凋亡和非凋亡途径的区分，对于细胞信号、与病原体的相互作用、治疗过程、药物发现、耐药性，甚至癌症和神经生成性疾病等疾病的发病机制的广泛应用至关重要。在这里，我们提出了一种基于无透镜数字全息术的新的高通量方法来识别细胞死亡过程中的凋亡性、坏死性和其他非凋亡性。这种方法依赖于识别哺乳动物细胞形态特征的时间变化，这是每个细胞死亡过程中独一无二的。已知的细胞毒剂可诱导不同的细胞死亡过程。基于深度学习的方法被用来自动分类细胞死亡机制(凋亡性、坏死性和非凋亡性)，准确率超过93%。这种无标签方法可以提供低成本(
Identification of cell death mechanisms, particularly distinguishing between apoptotic versus nonapoptotic pathways, is of paramount importance for a wide range of applications related to cell signaling, interaction with pathogens, therapeutic processes, drug discovery, drug resistance, and even pathogenesis of diseases like cancers and neurogenerative disease among others. Here, we present a novel high-throughput method of identifying apoptotic versus necrotic versus other nonapoptotic cell death processes, based on lensless digital holography. This method relies on identification of the temporal changes in the morphological features of mammalian cells, which are unique to each cell death processes. Different cell death processes were induced by known cytotoxic agents. A deep learning-based approach was used to automatically classify the cell death mechanism (apoptotic vs necrotic vs nonapoptotic) with more than 93% accuracy. This label free approach can provide a low cost (