How do you think the new GigE standards will influence the machine vision industry?
Respond or ask your question now!
Strictly speaking, the method is not wavelength dependent, but resolution is inversely related to wavelength. As a result, the team uses visible light rather than infrared, avoiding the ionizing radiation of the ultraviolet spectral region, which can damage samples. Switching from near-IR light to 532 nm light, for example, increased transverse resolution by a factor of two. The group has also experimented with a spatially coherent white-light source that eliminates laser speckle to yield better contrast. Using that source, they have been able to image single layers of graphene.
Their primary application focus right now, though, is medical diagnostics. For certain applications, such as assessing the refractive index and topography of red blood cells, the approach is ideal (see Figure 1). The cell mobility becomes a marker for their functionality and could be used in future for medical diagnostics.
The group seeks to refine the technology into a pathology instrument to complement existing instruments like cell counters and flow cytometers. Like other medical devices, it would need to be robust and accurate, but also high throughput, that is, to provide statistics over many thousands of cells in a matter of minutes. The software developed for the system can currently analyze around 2,500 cells in three seconds, which is impressive, but currently it takes the microscope 10 minutes to image those cells. Meanwhile, the size of the data sets presents another challenge. Run continuously, the system can produce more than a terabyte of data per day, which is impractical both in terms of storage and the amount of time required for processing. One approach that could address both the issues would be to analyze the entire field of view simultaneously rather than evaluating each cell individually.
At ASU, researchers have combined surface plasmon resonance and electrochemical detection techniques to develop electro-chemical current microscopy (ECCM), a method that can capture chemical data at high spatial resolutions. Surface plasmon resonance (SPR) imaging can provide high-spatial-resolution images of changes in local chemical concentrations, but without chemical specificity. Scanning electrochemical microscopy (SECM) provides chemical specificity but is slow and often invasive. Combining the two methods brings the best of both worlds. “SPR basically measures the change in the surface concentration of molecules and the electrochemical phenomenon always creates a change in the concentration of molecules,” says Tao, “so from that we can quantitatively determine electrochemical current from the SPR signal.” Instead of measuring current, the new microscopy measures current density, yielding high sensitivity, high-resolution results at video speed and with minimal perturbation to the measured electrochemical reactions.
In electrochemical detection, electrons transfer between the electrode and the sample under test, triggering a reduction-oxidation (redox) reaction and generating a characteristic electrochemical current for each material. Scanning microelectrodes over the sample under test, as in SECM, adds spatial resolution to the results, but it also imposes a constraint: Spatial resolution scales directly with microelectrode size but sensitivity scales with current level, which is inversely proportional to microelectrode size. Meanwhile, the scanning process can stretch data acquisition time to minutes, or even hours. In comparison, ECCM uses an area-array detector to capture data for the entire sample simultaneously. “We figured out a very simple way to determine electrochemical current from an optical signal,” says Tao.