A new method to eliminate image-degrading noise promises to boost the usefulness of optical coherence tomography (OCT)—and perhaps other modalities used for biomedical imaging.1
OCT works by splitting a light beam into two rays, one of which penetrates the specimen. A fraction of that light returns to its origin and interferes with the reference beam—and with other reflected light fields. "This adds ambiguities when interpreting the image," says Hon Luen Seck of the Singapore Institute of Manufacturing Technology at Agency for Science, Technology and Research (A*STAR, Singapore). Seck led the team that demonstrated reliable removal of the so-called autocorrelation noise by way of a reconstruction algorithm.
The approach works best with sparse samples that have relatively few features—so it's particularly appropriate for biological specimens. And besides OCT, it can be applied to other imaging methods. What's more, the researchers are working to speed up the reconstruction algorithm so the method can enable real-time applications. "We expect that with ongoing research the computational demand will decrease," Seck says.
1. H. L. Seck, Y. Zhang, and Y. C. Soh, J. Biomed. Opt., 17, 7, 076029 (2012).