Building a microfluidic device that captures cancer cells from blood for a liquid biopsy test







Independent research with Professor Wasserman and Manalis



Conducting a biopsy is a critical step in diagnosing cancer. Current techniques typically require a surgical intervention to collect tissue samples, a invasive procedure with high-risk. What if, instead, you could simply do a blood test?

The essential step of a "liquid biopsy" is the ability to capture the "circulating tumor cells" or CTC's from a blood sample. My challenge was to build such a device to capture these CTC's.

Sketch of the device that would capture CTCs from a genetically modified mouse in real time. A peristaltic pump would push the blood into the device, where the CTC's would be captured. The blood is then pumped back into the mouse. [Source: Manalis laboratory proposal]


Design Modules

I. Microfluidic Platform


The system to circulate blood through the device

II. Optical Detection System 


The system to detect and trigger the capture of a CTC in the device 

I. Microfluidic Platform

Using silicone micromachining and PDMS microfluidics I designed a device that would enable a suitable flow rate under pressure with parallel channels. 

2D rendering of microfluidic device built with soft-lithographic techniques 

AutoCad device design to extract CTCs from mouse blood. 

Critical design choices included: 

  • Choosing a push-up valve system because channels were deeper, allowing large particles in blood to flow through
  • Designing parallel flow channels to visualize a larger volume of blood
  • Strategically placing valves to enable switching speeds of the blood flow

These choices were made based on secondary research, expert interviews, prototyping and experimentation.


II. Optical Detection System

I designed and built a fluorescence microscope to detect when the CTC's pass through the microfluidic device. A visual trigger would automatically switch valves in the device to capture the CTC. 


2D sketch of the fluorescence microscope design


Critical design choices included:

  • Choosing a high numerical aperture, with a long-distance microscope objective to illuminate the underside of the device
  • Using a high brightness LED in-line with an excitation filter and a mirror as the illumination source


The device was successfully prototyped (after many-many tries!) and tested across multiple scenarios. The preliminary results were used by the Manalis lab to win the National Institute of Health RO1, a 5 year grant for cutting-edge research.