Page 59 - Vibrational Spectroscopic Imaging for Biomedical Applications
P. 59
36 Cha pte r T w o
diffraction patterns. Note that the first maximum is wavelength
dependent and is located at a larger diameter for longer wavelengths.
Importantly, the images clearly demonstrate the invariance of the
point-spread function at different positions on the sample plane as
suggested by Carr et al. 30
First measurements using IRENI are designed to show some of
the capabilities of the new instrument. A metal grid with grid bars of
the order of the wavelength is used as a test sample to show fre-
quency dependent behavior and a cluster of polystyrene beads of a
similar size is used to show an image due to a polystyrene absorp-
–1
tion band at 3025 cm . Figure 2.5 shows the transmission of light
through a metal grid with 8-μm-wide grid bars. Unprocessed images
–1
are shown at 3500, 2500, and 1500 cm . Note, as expected due to dif-
fraction, the images are blurrier at longer wavelengths. Above the
images, three spectra are shown from three individual 0.54 × 0.54 μm 2
pixels within the image. This entire dataset was collected within 1
minute. Figure 2.6 shows a visible and IR image of the absorption of
6-μm polystyrene beads. It is an unprocessed image corresponding
–1
to the absorption band at 3025 cm .
IRENI, a new facility at the SRC in Stoughton, Wisconsin, has
recently been commissioned, as demonstrated by the results pre-
sented above. As described below, these advances will make it pos-
sible to take time-resolved data on biological samples in vivo in the
near future.
To demonstrate the capabilities of the IRENI beam line and to
facilitate a direct comparison to state-of-the-art commercially avail-
able instruments utilizing global sources, the initial data was acquired
from a tissue biopsy of a benign prostate gland comprising of a layer
of compressed epithelial cells (Fig. 2.7). The three commercial instru-
19
ments that have been used to create images for direct comparison
include one that uses point mapping with an effective geometrical
area at the sample plane of 10 × 10 μm per pixel (Fig. 2.7a), one that
uses a linear array detector that detects from an area of 6.25 μm ×
6.25 μm per pixel (Fig. 2.7b) and one that uses a 64 × 64 pixel FPA
detector that detects from an area of 5.5 μm × 5.5 μm per pixel
(Fig. 2.7c). The IRENI beam line illuminates a 128 × 128 pixel FPA that
detects from an effective geometrical area at the sample plane of
0.54 μm × 0.54 μm per pixel (Fig. 2.7d).
2.3 Flow Cell for In Vivo IR Microspectroscopy
of Biological Samples
Fourier transform IR (FTIR) microspectroscopy’s capacity to nonde-
structively detect functional groups makes this technique a powerful
tool for studying biological specimen. 1–18 To date, however, the strong
absorption of liquid water in the mid-IR region as well as optical