Imaging with lensless microscopy sorts living cells
imec workers are applying nanoscale fabrication, lens-less optics, and micro-fluidics to the problem of identifying and sorting cells. The scenario is that of a small proportion of the cells circulating in the bloodstream being cancerous, and potentially capable of metastasising at new sites in the body. How to render them inactive?
next; lens-less microscopes…
A part of the solution is lens-less microscopy. The underlying principle is shown in this illustration. A sample is illuminated by a laser spot, and therefore by light with a planar wavefront. Under the sample (the same principles work in reflections, but this example uses transmission through the sample) is an light-sensitive array; the pixels in the array are illuminated by both the direct laser light, and by light diffracted by the sample.
Image processing can reconstruct a focussed, magnified image of the sample; the focal plane can be set by processing, and the field of view can be as large as the imaging chip. In the demonstration staged by imec, images were processed in real-time on a PC that was (loosely) described as having “good games machine” performance; a spokesman said that the processing was being carried out by code that includes some OpenCL programming of the graphics card’s GPU. With this ability to inspect individual cells, the “rogue” cells can be identified: imec proposes a real-time assessment in which the cells are passing through a micro-fluidics “gate” (pictured above, the cell-sorter) and in which those the microscope identifies as suspicious can be diverted and taken out of the main sample stream. The technique is termed “bubble-jet cell routing”.
In the nature of micro- and nano-fabrication, this is amenable to scaling and replication. The fluidic channels and the lens-less microscopy can be replicated many times over, in parallel, perhaps by as much as hundreds, or even a thousand instances; the microscopy algorithms can be coded into hardware and similarly parallelised. It becomes feasible, the researchers say, to think in terms of sorting 20million cells per second; at that speed, a quantity of blood could be taken from a patient, literally sorted to remove cancerous cells, and returned to the patient before it degraded.
Imec; www.imec.be
If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :
