Modeling and Control of Single-Cell Migration Induced by a Chemoattractant-Loaded Microbead

Abstract

Cell migration plays an essential role in cancer cell study. Investigation of a novel method for controlling cell migration movement can help develop new therapeutic strategies. In this paper, a chemoattractant-loaded microbead, which is controlled by optical tweezers, is used to stimulate a target cell to accomplish automated migration along a desired path while avoiding obstacles. Models of both tweezers-bead and bead-cell interactions are investigated. A dual closed-loop control strategy is proposed, which includes an inner tweezers-bead control loop and an outer bead-cell control loop. A proportional-integral feedback plus feedforward controller is used to control the inner loop, and an active disturbance rejection controller is used for the outer loop, which can address the cell migration modeling errors and unknown external disturbances. A traffic rule based on interference-clearing mechanism is also proposed to reduce external disturbances on the system by preventing other particles from interfering with the migration process. The effectiveness of the proposed control approach is verified by simulations and experiments on migrating leukemia cancer cells.

Active Disturbance Rejection Control of Euler–Lagrange Systems Exploiting Internal Damping

“…Specially, for underdamped systems, the gain matrix K d in (25) should be carefully set to avoid system instability, as the internal damping is lack in at least one channel. Sufficient external damping should be injected to the channels without internal damping.…”

“…(4) The last one is about the specific parameter setting of K d . By detailed analysis above, the gain K d in (25) for external damping injection can be determined following the subsequent three steps. That is: 1) estimate the damping inherent in the closed-loop system, i.e., M −1 D; 2) select the desired gain (25); 3) calculate the…”

“…where (14). For the sake of simplicity, all of the poles of the decoupled error dynamics (25) in three channels are placed at the same desired locations. It is given as follows:…”

“…It is important to note that, only very partial model information is employed in the typical ADRC framework, while the other known/unknown model information is regarded as the total disturbance. This framework has been widely applied to many practical systems, such as [16]- [25]. However, an important remark here is that the systems' physical structure has been neglected in this framework.…”

“…Here, it is needed to distinguish between M −1 Dė and Dq. M −1 Dė is the internal damping preserved in the closed-loop system(25), while Dq is the internal damping of the open-loop dynamic system…”