Xinyu Liu's Homepage

Xinyu Liu's Homepage

Whitesides Group

Harvard University

Home

Research Interests:

Ø Microrobotics

– Microrobotic manipulation of biomaterials (e.g., cells and organisms)

– Nanonewton force sensing and control by integrating MEMS devices

– Image processing and computer vision microscopy for cellular structure recognition and tracking

– Visual servoing with microscopy visual feedback

– Assimilation of multiple feedback modality (i.e., position, vision, and force) for intelligent microrobotic biomanipulation

– Microrobotic instrumentation for biomedical applications

Ø MEMS (microelectromechanical systems)

– Design, microfabrication, and testing of solid-state and polymeric MEMS devices

– BioMEMS for single cell manipulation and analysis

– MEMS-assisted nanorobotic manipulation and characterization of nanomaterials

Ph.D. Thesis Project: Autonomous Microrobotic Cell Manipulation

Video clip of fully-automated microrobotic zebrafish embryo injection (click here).

The main objective of this project is to develop microrobotic cell injection systems including fast sample immobilization for demonstrating fully automated, high-throughput, robust injection of zebrafish and mouse embryos with high survival, success, and phenotypic rates. The microrobotic systems will find important applications in genetic and reproductive research.

Based on automatic control and computer vision microscopy, a fully automated zebrafish embryo injection system has been developed enabling robust injection at a speed of 15 cells/min (compares favorable to the speed of manual operation) with high survival (98%, n=350), success (99%, n=350), and phenotypic (98.5%, n=210) rates. A vacuum-based embryo holding device is used for immobilizing a large number of embryos, reducing sample preparation from minutes to seconds. 810 zebrafish embryos were injected with fluorescent dyes and no-tail morpholino (ntl-MO) for quantitatively evaluating the system performance.

The manipulation strategies and control architecture, gained from the zebrafish embryo injection system, are used in my present research to develop a high-throughput microrobotic mouse embryo injection system, which will be applied to mitochondrial protein screening for assisted reproduction research.



Microrobotic zebrafish embryo injection system setup.



Vacuum based embryo holding device.	An array of immobilized embryos.



Recognition of zebrafish embryo structures. (A) After pre-processing. (B) Recognized chorion, cytoplasm center, switching point, and yolk/cell interface.



Embryos injected with fluorescent dye.	Ntl-MO injected fish (right) and uninjected control embryo (left).

Besides cell injection, this project also develops techniques to enable cellular force measurement and in situ mechanical characterization of individual cells during microinjection without requiring a separate process or system setup. In situ quantification of cellular mechanical properties during cell injection may prove that subtle mechanical differences are useful for embryo selection and health monitoring. In addition, quantification of cellular forces is also important in that cellular force feedback would enable force-controlled microrobotic cell manipulation and minimize injection-induced cell damage.

The cellular force measurement technique consists of a PDMS cell holding device and a sub-pixel visual tracking algorithm, which can be readily integrated into the cell injection system. Injection forces applied by a micropipette are transmitted to low-stiffness, protruding posts located inside a cavity, as shown in the schematic below. Post deflections, measured by the visual tracking algorithm, are fitted to an analytical mechanics model to obtain the injection force.

Importantly, the proposed cellular force measurement technique is not scale or cell type dependent. The PDMS cell holding device, constructed via soft lithography, can be readily modified to accommodate biological cells with different sizes. The technique has been applied to resolving cellular forces of both zebrafish embryos (~1.2mm) and mouse zygotes (~100µm) with nanoNewton measurement resolution (3.7nN). The experimental results demonstrate that force-deformation data can be used for mechanically distinguishing normal mouse embryos from those with compromised developmental competence (e.g., blastomere fragmentation).

In order to quantitate the mechanical properties of the injected/indented cells (e.g., Young’s modulus), a point-load elastic model of mouse oocyte/zygote is under development, which is capable of extracting the Young’s modulus of mouse oocyte/zygote from the measured cellular forces-deformation data.



Schematic of cellular force measurement.	SEM image of a PDMS holding device for mouse oocyte/zygote.



Injection force analysis. (a) Force balance on the cell under indentation. (b) Post deflection model.



A video clip showing tracking results of the post deflections (click here).	Young’s modulus calibration of the PDMS device.


Mouse zygotes for cellular force measurement. (a) Normal zygotes. (b) Zygote with blastomere fragmentation (arrow labeled).



Force-deformation curves of normal zygotes (blue) and fragmented zygotes (red). They separate themselves into two distinct regions.

Other projects:

During my Ph.D. studies, I also conducted several other projects in the fields of MEMS and microrobotics. Followings are representative illustrations for each project, and more details can be found in the corresponding research articles.

Ø A MEMS nanomanipulator with a sub-nanometer resolution


Solid model of the MEMS nanomanipulator with capacitive displacement sensor.



SEM picture of a released device.	Characterization results of the displacement sensor.

Ø A MEMS stage for 3-axis nanopositioning

SEM picture of a released MEMS nanopositioning stage.

Testing results of X, Y, and Z displacements vs. actuation voltage square.

Ø Real-time, high-accuracy micropipette aspiration for characterizing mechanical properties of biological cells



Schematic diagram of the micropipette aspiration system.	Visual tracking result of the cell deformation parameters.

Experimental results of interstitial cells (left) and human neutrophils (right) characterized by the micropipette aspiration system.

Ø Dynamic evaluation of autofocusing algorithms for automated microscopic analysis



Bright-field images of smear samples for evaluating autofocusing algorithms.


Dynamic focusing curves using Fibonacci search.	Focus curves of Variance algorithm that provides best overall performance.