Raman spectroscopy and a confocal microscope may enable medical laboratories to dramatically reduce the time it takes to identify microorganisms so that doctors can make the right diagnoses. Researchers at Erasmus University have demonstrated a spectroscopy method that can identify microorganisms after only six hours of growth on an agar plate. Because the technique is nondestructive, the samples can be analyzed with conventional means for added certainty. Conventional analysis takes one to three days, said Gerwin J. Puppels of the university's department of general surgery, and most of that time is spent obtaining pure cultures of microorganisms from the patient. Trimming this time, he said, speeds a doctor's ability to diagnose and then prescribe the appropriate treatment for a disease. "Different types of microorganisms will differ in their overall molecular composition and therefore have different Raman spectra," Puppels explained. "The Raman spectra can therefore serve as a sort of fingerprint." The researchers, including Puppels, recently published the results of a preliminary study on the technique in the Jan. 1, 2000, issue of Analytical Chemistry. To collect the spectra, the group used a Renishaw System 1000 Raman microspectrometer from Renishaw plc of Wotton-Under-Edge, UK. The system was attached to a Leica DM-LM microscope with an 80x objective. The researchers used a model 3900 Ti:sapphire laser operating at 830 nm from Spectra-Physics Lasers Inc. of Mountain View, Calif., to excite the spectra. Because microcolonies are only about 10 µm thick after six hours of culturing, the culture medium contributes to the collected signal. To remedy this, the group used vector correction. "In principle, it is very simple," Puppels said. "Suppose the spectra consist of 1000 datapoints. Then you can consider the spectra as vectors in a 1000-dimensional space." The result is that the microorganism vector will be orthogonal to the medium vector, no matter how much medium signal is present. Extending technology Puppels said that the project is part of a larger effort to develop techniques for rapid microorganism identification sponsored by the European Union's Biomed II program. It was an extension of confocal Raman microspectroscopy research with which he has been involved since the late 1980s. The work is "proof of principle," Puppels said. Much remains to be done before hospital laboratories will begin using Raman spectroscopy to test patient samples. He explained that there must be an extensive database of microorganisms' Raman spectra, and the ability of the technique to work at different sites with different instruments over time remains to be proved. He added that a universal, highly reproducible, synthetic growth medium would be beneficial.