448
H. M. Heise
best PLS calibration models led to SEP values between 0.30 and 0.52 mmol/l for the
removal of urea during hemodialysis sessions [21].
Just recently, an interesting study was published on urine analysis using a measurement system based on light-emitting diodes (LEDs) by considering future transition
to LED light sources. In their study, LEDs with ten standard wavelengths (1400–
2300 nm, in 100 nm increments [7150–4350 cm
-1 ]) were employed. The aim was
to estimate multiple components such as urea and creatinine in spot urine samples
[22]. A multiple regression analysis using all combinations of 10 wavelengths was
performed. NIR spectra from glucose-spiked urine samples from 10 healthy adults
were used for an appropriate wavelength selection. For estimating urinary urea and
creatinine levels, they obtained SEP values of 42 mg/dl and 7.34 mg/dl, respectively,
using four wavelengths for urea and five wavelengths for creatinine. This optical,
reagent-free method is suitable for practical determination of the urea-to-creatinine
ratio, which allows assessing protein intake in chronic kidney disease (CKD) patients.
A forensic application of NIR imaging is also presented, for which the identification and location of body fluid stains was investigated. For this application,
the potential of near-infrared hyperspectral imaging (NIR-HSI) was investigated
[23]. The authors used a hyperspectral camera (Specim, Oulu, Finland), which was
equipped with a mercury cadmium telluride (MCT) cryogenically cooled detector,
providing access to the spectral range from 1000–2500 nm [10,000–4000 cm
−1 ].
With the combination of the NIR-HSI data and simple chemometric techniques such
as principal component analysis (PCA) and classical least squares (CLS) regression,
the evidence and location of semen, vaginal fluid and urine in bodily fluid stained
fabrics could be revealed.
Another clinical application with an intended purpose for diabetes screening has
recently been reported. This indication is accessible due to the analysis of glycated
hemoglobin, especially of the HbA1c fraction, which is usually quantified by HPLC
methodology. An interesting review on optical methods for studying glycation of
proteins including hemoglobin, also with a focus on Raman spectroscopy, has been
given by Pandey et al. [24]. The HbA1c fraction is usually reported as percentage
against total hemoglobin (for healthy people < 6.0%). In Fig. 20.6, NIR spectra of
several proteins are given, which have been considered for glycation testing as there
are albumin, collagen, keratin and hemoglobin. Without blood testing, another assay
for a non-invasive assay for assessing the glycation rate in people with diabetes has
been suggested based on glycated keratin as available from finger nail clippings [25].
For testing their hypothesis, the researchers used artificial non-enzymatic glycation
of keratin in aqueous glucose solution. For the in vitro glycation samples, spectral
changes versus the spectra of untreated keratin were observed in the range between
4300 and 7500 cm
−1 . Spectral classification based on people with diabetes mellitus
versus healthy subjects was performed using partial least square discriminant analysis (PLS-DA). Using standard normal variate normalization and Savitzky-Golay
smoothing with first derivative preprocessing, a prediction rate with 100% correct
assignments was achieved for a test set. This result is seen with some skepticism based
on our study of stratum corneum with spectra measured non-invasively by attenuated
total reflection mid-infrared spectroscopy. The data base included healthy subjects
H. M. Heise
best PLS calibration models led to SEP values between 0.30 and 0.52 mmol/l for the
removal of urea during hemodialysis sessions [21].
Just recently, an interesting study was published on urine analysis using a measurement system based on light-emitting diodes (LEDs) by considering future transition
to LED light sources. In their study, LEDs with ten standard wavelengths (1400–
2300 nm, in 100 nm increments [7150–4350 cm
-1 ]) were employed. The aim was
to estimate multiple components such as urea and creatinine in spot urine samples
[22]. A multiple regression analysis using all combinations of 10 wavelengths was
performed. NIR spectra from glucose-spiked urine samples from 10 healthy adults
were used for an appropriate wavelength selection. For estimating urinary urea and
creatinine levels, they obtained SEP values of 42 mg/dl and 7.34 mg/dl, respectively,
using four wavelengths for urea and five wavelengths for creatinine. This optical,
reagent-free method is suitable for practical determination of the urea-to-creatinine
ratio, which allows assessing protein intake in chronic kidney disease (CKD) patients.
A forensic application of NIR imaging is also presented, for which the identification and location of body fluid stains was investigated. For this application,
the potential of near-infrared hyperspectral imaging (NIR-HSI) was investigated
[23]. The authors used a hyperspectral camera (Specim, Oulu, Finland), which was
equipped with a mercury cadmium telluride (MCT) cryogenically cooled detector,
providing access to the spectral range from 1000–2500 nm [10,000–4000 cm
−1 ].
With the combination of the NIR-HSI data and simple chemometric techniques such
as principal component analysis (PCA) and classical least squares (CLS) regression,
the evidence and location of semen, vaginal fluid and urine in bodily fluid stained
fabrics could be revealed.
Another clinical application with an intended purpose for diabetes screening has
recently been reported. This indication is accessible due to the analysis of glycated
hemoglobin, especially of the HbA1c fraction, which is usually quantified by HPLC
methodology. An interesting review on optical methods for studying glycation of
proteins including hemoglobin, also with a focus on Raman spectroscopy, has been
given by Pandey et al. [24]. The HbA1c fraction is usually reported as percentage
against total hemoglobin (for healthy people < 6.0%). In Fig. 20.6, NIR spectra of
several proteins are given, which have been considered for glycation testing as there
are albumin, collagen, keratin and hemoglobin. Without blood testing, another assay
for a non-invasive assay for assessing the glycation rate in people with diabetes has
been suggested based on glycated keratin as available from finger nail clippings [25].
For testing their hypothesis, the researchers used artificial non-enzymatic glycation
of keratin in aqueous glucose solution. For the in vitro glycation samples, spectral
changes versus the spectra of untreated keratin were observed in the range between
4300 and 7500 cm
−1 . Spectral classification based on people with diabetes mellitus
versus healthy subjects was performed using partial least square discriminant analysis (PLS-DA). Using standard normal variate normalization and Savitzky-Golay
smoothing with first derivative preprocessing, a prediction rate with 100% correct
assignments was achieved for a test set. This result is seen with some skepticism based
on our study of stratum corneum with spectra measured non-invasively by attenuated
total reflection mid-infrared spectroscopy. The data base included healthy subjects