Application of Resonance Raman Spectroscopy to the Detection of Carotenoids In Vivo - Carotenoids: Physical, Chemical, and Biological Functions and Properties - page 96

CCD camera

L3

F3
F2

F5

F4

BS2

Aiming
beam

F1

CL

Fiber

BS1

Shutter

L1

AP

L2

F4

F1

BS
AP
L3

Living eye

L1

L2

Excised eye

(a) (b)

FIGURE 6.7 (a) Schematics of experimental setup used for in vivo resonance Raman imaging, RRI, of MP
distributions. Light from a blue laser source is projected onto the macula as a ~3.5 mm diameter excitation
disk. The backscattered light is collimated by the lens of the eye and imaged with a two-dimensional CCD
camera array detector. Two sets of i lters are used sequentially to selectively image light at the C = C Raman
wavelength (Raman image) and at a slightly longer wavelength (offset image). The two images are digitally
subtracted and displayed as topographic or three-dimensional pseudocolor images of the spatial MP concen-
trations. L1-3: lenses; F1: laser line i lter; BS: dichroic beam splitters; F2: tunable i lter; and F3: band pass
i lter. Inset shows modii cations for use with excised tissue. (b) Photograph of subject measured with instru-
ment. RRI images are recorded with 0.2 s exposure time for dilated or non-dilated pupils.

light returned from the retina is i ltered to only transmit l uorescence components slightly above the
Raman wavelength, at λ offset .

The contribution of the broad l uorescence at the Raman wavelength λ R is approximately the

same as at the slightly offset longer wavelength position λ offset . It can be shown, Equation 6.2, that

the Raman component I R ( λ R ) for each image pixel is approximately

()

− 1 − 1

(

() ( )

)

(6.2) I λ≈ T ( λ ⋅ ) T ( λ ) I λ T − I λ T

R R OM exc OM R Det R R Det offset offset

where
I Det ( λ offset ) are the detector intensities

T R and T offset are the i lter transmissions at the respective wavelengths
T OM is the unknown transmission of the ocular media

λ R ) and I Det (

The RRI image of an MP distribution can thus be derived with a digital image subtraction routine,
where the intensities obtained for each pixel of the two images are divided by the appropriate i lter
transmission coefi cient.