465
15 FtIR Imaging of tissues: techniques and methods of Analysis
15.6 Outlook
the scene is now set for FtIR to play a significant role in pathological diagnosis;
however, a number of hurdles need to be overcome prior to implementation into the
clinical laboratory. one of the most important barriers to overcome is standardization of methodology, instrumental procedures and data processing. At present every
lab working on FtIR imaging of tissue sections have their own method of sample
preparation, instrumental parameters and spectral processing. these methods need
to be standardized so a measurement recorded in Australia on one section would
yield a similar result to the same section prepared and measured in Poland, for
example.
the inability to detect strong contributions from A-dNA in fixed tissue sections
is problematic given that the importance of dNA in the etiology of the disease. In
the future it may be possible to fix B-dNA and resolve the important phosphodiester vibrations and other dNA marker bands. At present the only way to preserve
B-dNA is to keep the sample in a hydrated state, which is clearly not conducive for
routine diagnosis.
the contribution of light dispersion and mie scattering to the spectral profile
needs more investigation particular on how this effect influences the histological
diagnosis. the question at present is do we correct for this light dispersion by performing Rmie-EmSC (or some other approach) or do we leave the mie scattering
component in the spectrum because it too is diagnostically useful. If we consider
that the size of the nucleus is one of the most important pathological criteria that a
histologist uses to grade cancer then it seems counterintuitive that one would want
to remove this information from the spectrum. the same argument can be made
for dispersion. the optically dense regions of tissue are often the cancerous areas
containing abundant nuclei. It is clear from the early studies applying FtIR imaging
to diagnose CIN in cervical tissues that when the amide I and II region was used in
the cluster analysis the best demarcation of cancer versus non-cancer was achieved
[31]. this region of course is the most affected by dispersion as this is the region of
highest absorbance in the FtIR spectrum. As biospectroscopists we try to achieve
a “text book” spectrum devoid of baseline undulations and other physical contributions in pursuit of the pure chemical information, however, if our purpose is to develop diagnostic tools then one could argue that all contributions whether physical
or chemical should be included in the ensuing analysis. the focus of future studies
should address this central and pivotal question.
the other important consideration is choice of substrate given the inherent standing-wave-effect generated by IR trans-reflection slides. this effect can be reduced,
in the case of tissue, if each section is evenly cut and a second derivative calculated
15 FtIR Imaging of tissues: techniques and methods of Analysis
15.6 Outlook
the scene is now set for FtIR to play a significant role in pathological diagnosis;
however, a number of hurdles need to be overcome prior to implementation into the
clinical laboratory. one of the most important barriers to overcome is standardization of methodology, instrumental procedures and data processing. At present every
lab working on FtIR imaging of tissue sections have their own method of sample
preparation, instrumental parameters and spectral processing. these methods need
to be standardized so a measurement recorded in Australia on one section would
yield a similar result to the same section prepared and measured in Poland, for
example.
the inability to detect strong contributions from A-dNA in fixed tissue sections
is problematic given that the importance of dNA in the etiology of the disease. In
the future it may be possible to fix B-dNA and resolve the important phosphodiester vibrations and other dNA marker bands. At present the only way to preserve
B-dNA is to keep the sample in a hydrated state, which is clearly not conducive for
routine diagnosis.
the contribution of light dispersion and mie scattering to the spectral profile
needs more investigation particular on how this effect influences the histological
diagnosis. the question at present is do we correct for this light dispersion by performing Rmie-EmSC (or some other approach) or do we leave the mie scattering
component in the spectrum because it too is diagnostically useful. If we consider
that the size of the nucleus is one of the most important pathological criteria that a
histologist uses to grade cancer then it seems counterintuitive that one would want
to remove this information from the spectrum. the same argument can be made
for dispersion. the optically dense regions of tissue are often the cancerous areas
containing abundant nuclei. It is clear from the early studies applying FtIR imaging
to diagnose CIN in cervical tissues that when the amide I and II region was used in
the cluster analysis the best demarcation of cancer versus non-cancer was achieved
[31]. this region of course is the most affected by dispersion as this is the region of
highest absorbance in the FtIR spectrum. As biospectroscopists we try to achieve
a “text book” spectrum devoid of baseline undulations and other physical contributions in pursuit of the pure chemical information, however, if our purpose is to develop diagnostic tools then one could argue that all contributions whether physical
or chemical should be included in the ensuing analysis. the focus of future studies
should address this central and pivotal question.
the other important consideration is choice of substrate given the inherent standing-wave-effect generated by IR trans-reflection slides. this effect can be reduced,
in the case of tissue, if each section is evenly cut and a second derivative calculated