In vivo confocal Raman spectroscopy of the human cornea

Purpose. To investigate the feasibility of a confocal Raman spectroscopic t echnique (see ref. 13) for the noninvasive assessment of corneal hydration in vivo in two legally blind subjects. Methods. A laser beam (632.8 nm; 15 mJ) was maintained on the cornea by using a microscope objective lens (x25 magnification, NA = 0.5, f = 10 mm) both for focusing the incident light as well as collecting the Raman backscattered light, in a 180 degrees backsca tter configuration. An optical fiber, acting as the confocal pinhole for el imination of light from out-of-focus places, was coupled to a spectrometer that dispersed the collected light onto a sensitive array detector for rapi d spectral data acquisition over a range from 2,890 to 3.590/cm(-1) Raman s pectra were recorded from the anterior 100-150 mu m of the cornea over a pe riod before and after topical application of a mild dehydrating solution. T he ratio between the amplitudes of the signals at 3,400/cm(-1) (OH-vibratio nal mode of water) and 2,940/cm(-1) (CH-vibrational mode of proteins) was u sed as a measure for corneal hydration. Results, High signal-to-noise ratio (SNR = 25) Raman spectra were obtained from the human corneas by using 15 mJ of laser light energy. Qualitative changes in the hydration of the anter iormost part of the corneas could be observed as a result of the dehydratin g agent. Conclusion, With adequate improvements in system safety, confocal Raman spectroscopy could potentially be applied clinically as a noninvasive tool for the assessment of corneal hydration in vivo.