期刊
BIOSENSORS & BIOELECTRONICS
卷 133, 期 -, 页码 16-23出版社
ELSEVIER ADVANCED TECHNOLOGY
DOI: 10.1016/j.bios.2019.03.002
关键词
Biophysical cellular biomarker; Cell deformability; Electrical impedance spectroscopy; Microfluidic cytometry; Single cell analysis
类别
资金
- Singapore-MIT Alliance for Research and Technology (SMART) Centre, BioSystems and Micromechanics (BioSyM) IRG - National Research Foundation (NRF) of Singapore
- Singapore Ministry of Education - Singapore Academic Research Fund Tier 2 [T2MOE1603]
- Ministry of Education (MOE), Singapore
- BioSyM IRG
Precise measurement of mechanical and electrical properties of single cells can yield useful information on the physiological and pathological state of cells. In this work, we develop a differential multiconstriction microfluidic device with self-aligned 3D electrodes to simultaneously characterize the deformability, electrical impedance and relaxation index of single cells at a high throughput manner ( > 430 cell/min). Cells are pressure driven to flow through a series of sequential microfluidic constrictions, during which deformability, electrical impedance and relaxation index of single cells are extracted simultaneously from impedance spectroscopy measurements. Mechanical and electrical phenotyping of untreated, Cytochalasin B treated and N-Ethylmaleimide treated MCF-7 breast cancer cells demonstrate the ability of our system to distinguish different cell populations purely based on these biophysical properties. In addition, we quantify the classification of different cell types using a back propagation neural network. The trained neural network yields the classification accuracy of 87.8% (electrical impedance), 70.1% (deformability), 42.7% (relaxation index) and 93.3% (combination of electrical impedance, deformability and relaxation index) with high sensitivity (93.3%) and specificity (93.3%) for the test group. Furthermore, we have demonstrated the cell classification of a cell mixture using the presented biophysical phenotyping technique with the trained neural network, which is in quantitative agreement with the flow cytometric analysis using fluorescent labels. The developed concurrent electrical and mechanical phenotyping provide great potential for high-throughput and label-free single cell analysis.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据