Page 288 - Optofluidics Fundamentals, Devices, and Applications

262 Cha pte r Ele v e n

translation of the object across the grid (Fig. 11-1c). As the object passes
across each hole, the time-varying transmission represents a line scan
across the object. By choosing the small angle between the grid orienta-
tion and the translation direction, we can ensure that the object is fully
scanned by the apertures. This design can be further simplified by
replacing the tilted 2D aperture grid with a long tilted 1D aperture
array (Fig. 11-1d). This imaging strategy is the basis of the optofluidic
microscopy (OFM) method. The OFM method shares a lot of similari-
ties with near-field scanning optical microscopy methods. In fact, the
OFM aperture array can be interpreted as a series of NSOM apertures.
Whereas NSOM sensors are generally raster-scanned over the target
objects, the OFM approach uses object translation to accomplish scan-
ning. This is a significant advantage for objects that are suspended in
fluids as we can apply microfluidic technology to implement flow con-
trols in a compact and cost-effective fashion.
In terms of implementation, our current typical OFM prototype
consists of a metal-coated sensor with apertures etched onto the metal
layer as the base layer. The top layer consists of a transparent struc-
ture containing a carefully aligned microfluidic channel for sample
delivery and scanning. An illumination source situated above the
device completes the design. To perform imaging, we flow the targets
through the channel and electronically acquire line scans. The image
composition processing is minimal and simply involves compiling
the line scans appropriately.

11-3 Prototype Evaluations

11-3-1 Caenorhabditis elegans Imaging
Our on-chip OFM prototype (Fig. 11-2a and 11-2b) utilizes the above-
mentioned core design with one change—two parallel OFM arrays
are implemented (Fig. 11-2c). We choose to use two parallel OFM
arrays for two reasons. First, by measuring the time difference
between when the target object first passes across each array, we can
determine the flow speed of the object by dividing the distance sepa-
ration between the arrays by that time difference. Knowledge of the
speed is important for the correct computation of the delay and the
correct matching of the collected line scans to generate OFM images.
Second, significant differences in the two acquired images will indi-
cate object shape changes, flow speed variations, and/or object rota-
tions during the data-acquisition process. Accurate OFM imaging
requires the absence of these variations, and therefore, discrepancy in
the images is a good criterion for rejecting that image pair. In our
experiments, we reject image pairs when their correlation is less than
50%. During our initial experiments, approximately 50% of the sam-
ples were rejected based on this criterion.