Automated estimation of cancer cell deformability with machine learning and acoustic trapping

Abstract

Cell deformability is a useful feature for diagnosing various diseases (e.g., the invasiveness of cancer cells). Existing methods commonly inflict pressure on cells and observe changes in cell areas, diameters, or thickness according to the degree of pressure. Then, the Young’s moduli (i.e., a measure of deformability) of cells are estimated based on the assumption that the degrees of the changes are inversely proportional to Young’s moduli. However, manual measurements of the physical changes in cells are labor-intensive, and the subjectivity of the operators can intervene during this step, thereby causing considerable uncertainty. Further, because the shapes of cells are nonuniform, we cannot ensure the assumption for linear correlations of physical changes in cells with their deformability. Therefore, this study aims at measuring non-linear elastic moduli of live cells (degrees of cell deformability) automatically by employing conventional neural networks (CNN) and multilayer perceptrons (MLP) while preserving (or enhancing) the accuracy of the manual methods. First, we obtain photomicrographs of cells on multiple pressure levels using single-beam acoustic tweezers, and then, we suggest an image preprocessing method for emphasizing changes in cell areas on the photomicrographs. The CNN model is trained to measure the ratios of the cell area change at each pressure level. Then, we apply the multilayer perceptron (MLP) to learn the correlations of the cell area change ratios according to the pressure levels with cell deformability. The accuracy of the CNN was evaluated using two types of breast cancer cells: MDA-MB-231 (invasive) and MCF-7 (noninvasive). The MLP was assessed using five different beads (Young’s moduli from 0.214 to 9.235 kPa), which provides standardized reference data of the non-linear elastic moduli of live cells. Finally, we validated the practicality of the proposed system by examining whether the non-linear elastic moduli estimated by the proposed system can distinguish invasive breast cancer cells from noninvasive ones.

Critical review of single-cell mechanotyping approaches for biomedical applications

“…Noting that cells trapped in the Rayleigh regime also experience compression in the direction of beam propagation, later studies applied this technique for the characterisation of cell deformability by analysing the projected area of a trapped cell in the plane perpendicular to the compression. 77,177,178,189 SBAT devices applied in this way have been used to study cancer cells, finding similar trends of increasing deformability with invasiveness as other techniques, 177,178 while also supporting a novel use case for targeted cancer cell destruction using high acoustic powers. 177 Recent advances in SBAT technology include calibration of deforming forces via comparative micropipette aspiration measurements 177 and the application of a deep learning model to automatically estimate nonlinear elastic moduli of live cells.…”

“…177 Recent advances in SBAT technology include calibration of deforming forces via comparative micropipette aspiration measurements 177 and the application of a deep learning model to automatically estimate nonlinear elastic moduli of live cells. 77 Collectively, acoustofluidic deformation mechanotyping techniques have been applied to red blood cells, 49,52 breast cancer cells, 77,177 acute lymphoblastic leukemia cells 189 and select tissue cells. 186 However, each of these studies were restricted to measurement timescales on the order of one minute per cell, and further work is required to prove its efficacy in a high-throughput regime.…”

“…In doing so we highlight developing approaches that combine low fabrication cost, microfluidic cell handling and the ability to modulate deforming forces, 52,75 in addition to recent developments supporting the potential for machine learning to enhance characterization. 76–80 We also identify directions for future work to improve the throughput, data quality and reliability of single-cell mechanotyping measurements.…”

A Review of Single-Cell Microrobots: Classification, Driving Methods and Applications

“…Hwang et al [ 255 ] used acoustic tweezers to precisely attach fibronectin-coated microbeads to human breast cancer cells and then stretched the cell membrane by remotely pulling the cell-attached microbeads with an acoustic trap and found that highly invasive breast cancer cells exhibited greater deformability than weakly invasive breast tumor cells. Recently, Lee et al [ 256 ] used an artificial neural network (CNN) to measure the rate of cell area change at each pressure level after deforming the cells using acoustic tweezers and then applied a multilayer perceptron (MLP) to learn the correlation between the rate of cell area change according to the pressure level and the deformability of the cells. The system was trained to discriminate against invasive breast cancer cells and noninvasive breast tumor cells.…”

Ultrasonic High-Resolution Imaging and Acoustic Tweezers Using Ultrahigh Frequency Transducer: Integrative Single-Cell Analysis

“…Acoustic tweezers are an emerging tool in the field of biomedical and physical research [ 45 , 46 , 47 , 48 ] due to their non-contact and strong radiation forces [ 49 , 50 ]. They have been widely used for single-cell analysis, such as measuring calcium elevation and mechanical properties of cells [ 51 , 52 , 53 , 54 , 55 ]. By integrating the capabilities of both cell imaging and acoustic tweezer into a single UHF transducer, we demonstrate the feasibility of performing non-invasive trapping and high-resolution imaging of single cells and organelles in a dual mode.…”