The present invention relates to a method for providing information on the presence of viral infection by measuring the Raman spectrum of a dried tear sample prepared on a substrate and by extracting multiple Gaussian peaks therefrom to evaluate the intensity ratio of two specific wavelengths; and to a diagnostic device for viral infection using the same.
Currently, for the diagnosis of infectious diseases, methods comprising multiple steps of collecting and culturing cells, collecting genes therefrom, and amplifying the genes by polymerase chain reaction (PCR) to confirm, are used. While these methods require much time and effort, it is important for such infectious diseases to be quickly diagnosed and properly treated because they are contagious in many cases. Therefore, a method for quickly and easily diagnosing infectious disease is required.
Tear analysis methods based on Raman spectroscopy have recently been studied for the research on infectious ocular surface diseases. For example, Korean Patent No. 10-1336478 discloses detection of viral particles in tear films using surface-enhanced Raman spectroscopy (SERS).
However, in the case of the tear analysis methods based on the Raman spectroscopy of prior art, they diagnose the presence or absence of a virus in a sample by analyzing the difference of overall SERS spectrum patterns, it is difficult to analyze the difference because they compare the entire spectrum patterns, and there is a problem in that the boundary for accurately diagnosing an infection is not clear.
It is an object of the present invention to provide a method and device for diagnosing viral infections, which are derived from the technical background described above, and which can diagnose infectious diseases quickly and simply.
Another object of the present invention is to provide a stand-alone diagnostic device for viral infection, which can be used for the diagnosis of infectious diseases at clinical sites; and a method for diagnosing viral infection using the same.
Another object of the present invention is to provide a method and device for diagnosing viral infection, in which the presence of viral infection can be accurately diagnosed by tear analysis methods based on Raman spectroscopy.
The present invention provides a method for providing information on the presence of viral infection, comprising: a first step of preparing a dried tear sample on a substrate; a second step of measuring a Raman spectrum from the dried tear sample; a third step of extracting Gaussian sub-peaks by deconvolution of the measured Raman spectrum; a fourth step of deriving a log value for the relative intensity ratio of a peak corresponding to an amide III β-sheet and a peak corresponding to C—H deformation by Equation 1 below; and a fifth step of determining the sample as normal if the derived value is positive and as infected if the derived value is negative:
Further, the present invention provides a diagnostic device for viral infection, comprising: a detection substrate capable of providing a dried tear sample by applying a teardrop thereon; a signal measuring unit for measuring a Raman signal from the inserted detection substrate; a peak deconvolution unit for separating the measured Raman peaks into Gaussian sub-peaks; a data processing unit for deriving a log value for a relative ratio of two peaks appearing at specific wavelengths among the separated Gaussian sub-peaks; and a display unit for showing the derived value.
According to the present invention, about 10 overlapped peaks appearing in a range of 1200 cm−1 to 1500 cm−1 are separated into single Gaussian peaks from the spectrum obtained using drop-coating deposition surface-enhanced Raman spectroscopy (DCD-SERS) in which surface-enhanced Raman scattering and drop-coating deposition are fused, and the relative intensity ratio of two specific peaks therefrom, particularly, peaks appearing at 1342 cm−1 and 1242 cm−1, can be evaluated to confirm the presence of adenoviral infection.
The method for diagnosing viral infection of the present invention can diagnose viral infection faster than conventional PCR methods, and the present invention can provide a stand-alone diagnostic device which can be used for the diagnosis of infectious diseases at clinical sites.
The present invention provides a method for providing information on the presence of viral infection, comprising: a first step of preparing a dried tear sample on a substrate; a second step of measuring a Raman spectrum from the dried tear sample; a third step of extracting Gaussian sub-peaks by deconvolution of the measured Raman spectrum; a fourth step of deriving a log value for the relative intensity ratio of a peak corresponding to an amide III β-sheet and a peak corresponding to C—H deformation by Equation 1 below; and a fifth step of determining the sample as normal if the derived value is positive and as infected if the derived value is negative:
The present invention is based on the first finding that viral infection can be diagnosed by evaluating the peak intensity ratio at two specific wavelengths by deconvoluting the Raman spectrum for dried tear samples, in which about 10 Gaussian peaks appear by being overlapped, into individual Gaussian peaks. For example, in the case of patients suffering from conjunctivitis due to adenoviral infection, by confirming that a log value of the relative ratio of the peak intensity at 1242 cm−1 to the peak intensity at 1342 cm−1 changed from a positive value to a negative value, these two peaks were identified as useful parameters for the diagnosis of adenoviral infection, and a method for diagnosing infection using the same was suggested.
Preferably, the first step may be performed by drop-coating deposition (DCD).
Preferably, the second step may be performed by surface-enhanced Raman spectroscopy.
Preferably, the substrate may be a support coated with nanoparticles. By using a nanoparticle-coated support, the sensitivity of measurements can be improved by inducing surface-enhanced Raman scattering. In general, Raman scattering is excellent in selectivity, but has a disadvantage in that detection is not easy due to weak signal intensity as compared with other optical detection methods such as absorption, fluorescence, etc. Therefore, in order to overcome this, it is necessary to use a highly sensitive detector, or a method capable of increasing the signals is needed. Accordingly, by using a support coated with nanoparticles, Raman signals generated by the surface enhancement effect due to the nanoparticles can be enhanced, and thus measurements can be performed without the aid of a special detector.
Preferably, the measurements may be performed at the central (C) zone, middle (M) zone, or secondary ring (T) zone of the dried tear sample.
In a specific exemplary embodiment of the present invention, as a result of measuring and analyzing the Raman spectra at the four zones of the dried tear samples, namely, the C, M, T, and R zones, a significant change in relative signal intensity was observed at two selected wavelengths at C, M, and T zones, but the change observed at the R zone was negligible (
Preferably, the peak corresponding to the amide III β-sheet may appear in a range of 1242±10 cm−1, and the peak corresponding to C—H deformation may appear in a range of 1342±10 cm−1, respectively.
Preferably, the method for providing information of the present invention may provide information on the presence of adenoviral infection.
In a specific exemplary embodiment of the present invention, tear samples from adenoviral conjunctivitis-confirmed patients and from non-infected persons were compared, and as a result, it was confirmed that in the Raman spectrum of the non-infected samples, the log value of the intensity ratio of the peak at 1242 cm−1 corresponding to the amide III β-sheet to the peak at 1342 cm−1 corresponding to C—H deformation was always positive, but in the spectrum of adenoviral conjunctivitis-confirmed patients, the ratio was remarkably decreased, showing a negative log value. That is, a spectrum appearing by about 10 overlapped peaks in the range of 1200 cm−1 to 1500 cm−1 were resolved into single Gaussian peaks, and by evaluating the relative intensity ratio of two specific peaks among those peaks, in particular, the peaks appearing at 1342 cm−1 and 1242 cm−1, it was confirmed that it was possible to determine the presence of adenoviral infection.
Further, the present invention provides a diagnostic device for viral infection, comprising: a detection substrate capable of providing a dried tear sample by applying a teardrop thereon; a signal measuring unit for measuring a Raman signal from the inserted detection substrate; a peak deconvolution unit for separating the measured Raman peaks into Gaussian sub-peaks; a data processing unit for deriving a log value for a relative ratio of two peaks appearing at specific wavelengths among the separated Gaussian sub-peaks; and a display unit for showing the derived value.
Preferably, the diagnostic device of the present invention may further comprise an input unit into which the detection substrate is inserted.
Preferably, in the diagnostic device of the present invention, the signal measuring unit may comprise a light source and photon detector, and optionally further comprise a mirror, lens, and a filter.
Hereinbelow, the constitution and effects of the present invention will be described in detail with accompanying exemplary embodiments. However, the exemplary embodiments disclosed herein are only for illustrative purposes and should not be construed as limiting the scope of the present invention.
Among patients who visited Kyung Hee University Hospital, tear samples were collected from 8 patients (36±14 yr) who had been confirmed with adenoviral conjunctivitis and 8 normal persons (33±8 yr) with their consent. The present study has passed the IRB KMCIRN1401-02 at Kyung Hee University.
Tear collection was performed for 5 minutes at a nasoinferior conjunctival sac using a polyester-fiber rod (Transorb Wick, USA) with a diameter of 4 mm and a length of 10 mm without external stimulation. The rod which was removed from the eye was placed in an Eppendorf tube and centrifuged at 8,000 rpm for 15 minutes to remove the rod, and thereafter, it was sealed with parafilm (Pechiney, Plastic Packing Company, USA) and stored at −70° C. for 24 hours. It did not exceed 24 hours until the measurement was performed according to the present invention.
In order to obtain Raman spectra from the collected tear samples, a DCD-SERS spectral method was used, in which surface-enhanced Raman scattering (SERS) and drop-coating deposition (DCD) are fused. Specifically, a 50 nm Au-coated anodized aluminum oxide (AAO) nanodot array substrate and a commercially available 2.5 nm Ti- and 50 nm Au-coated Au.0500.ALSI (Platypus Technologies, USA) substrate were used. Approximately 2 μL of tear was dropped on a clean substrate and dried to prepare samples for measurement. A SENTERRA confocal Raman system (Bruker Optics Inc., USA) equipped with a 785 nm diode laser source with a 200 mW output was used. In addition, it was possible to measure with a portable Raman. The examination was performed for 30 seconds by laser-irradiating the dried tear, which was sectioned into four zones (C, M, T, and R zones, respectively, from the center), according to known methods. The measured spectra were in a range of 417 cm−1 to 1782 cm−1, and the central spectrum was 1200 cm−1.
2.1. Diagnostic Marker 1: AC
As shown in the Equation below, the log value of the ratio of the Raman intensity at a wavelength of 1242 cm−1 corresponding to an amide III β-sheet relative to the Raman intensity at a wavelength of 1342 cm−1 corresponding to C—H deformation was defined as an AC biomarker (see the Equation below). Whereas in non-infected normal tear, the amide III β-sheet at a wavelength of 1242 cm−1 always had a greater value than the C—H deformation at a wavelength of 1342 cm−1 so that the AC diagnostic marker always showed a positive value, in the case of conjunctivitis patients infected with adenovirus, the relative intensity of the peak at 1342 cm−1 was increased, and the AC marker showed a negative value.
In the Equation above, I1242 and I1342 are Raman intensities at wavelengths of 1242 cm−1 and 1342 cm−1, respectively. The above calculation was performed using MATLAB software.
2.2 Diagnostic Marker 2: Principal Component Analysis (PCA) Algorithm
Principal component analysis is a data processing technique that is useful for visualization and feature extraction of data, as well as dimensional reduction of feature vectors for reducing high dimensional feature vectors to low dimensional feature vectors. Three DCD-SERS spectra, each defined at 1242 cm−1, 1342 cm−1, and 1448 cm−1, were used as inputs for a transfer function to detect the presence of adenoviral infection. Specifically, three vectors [1242 cm−1, 1342 cm−1], [1242 cm−1, 1448 cm−1], and [1342 cm−1, 1448 cm−1], which were normalized by the Z-score method, were used as inputs for the proposed transfer function. The performance of the principal component analysis was evaluated by the receiver operating characteristic curve (ROC curve) analysis, and the algorithm therefor was implemented in MATLAB software.
2.3 Diagnostic Marker 3: Deconvolution Method for Multiple Gaussian Peaks (MGPs)
In order to distinguish the difference between the normal condition and conjunctivitis due to adenoviral infection, a method for resolving multiple Gaussian peaks from the DCD-SERS spectrum was used. That is, the discrete version of a single Gaussian function can be defined by the Equation below:
In the Equation above, Hk is the amplitude of the single Gaussian function, fk is a maximum frequency position of the single Gaussian function, and wk is a half-width of the single Gaussian function.
The Gaussian curve of the optimized spectrum by using the above Equation can be expressed as the sum of Gaussian functions as shown by the Equation below.
In the Equation above, m is the total number of Gaussian functions.
The DCD-SERS spectrum in the range of 1200 cm−1 to 1500 cm−1 was used as the input for the multi-Gaussian model for feature peak extraction using the above equation. In order to extract multiple Gaussian peaks (MGPs) from the measured spectrum, m=10, that is, 10 Gaussian peaks were selected to have 30 cm−1 wavelength intervals within the range. From the four zones of a dried teardrop, wavelength shift (Raman shift), amplitude (Raman intensity), half-width, and area of Gaussian peaks were extracted and evaluated. An algorithm for extracting multiple Gaussian feature peaks using Gaussian resolution was also implemented in MATLAB software.
In order to compare the differences in mean values between two groups, for statistical analysis with a basic expression of mean and standard deviation, two-tailed Student's t-test method was used, and the intensity of the morphological DCD-SERS spectrum of the dried teardrop was analyzed using one-way analysis of variance (ANOVA). The Student-Newman-Keuls test was used for post hoc comparison. In order to evaluate the analytical efficiency of the AC biomarker, clinical analyses such as sensitivity, specificity, accuracy, prevalence, and error rates were used, and in order to evaluate the efficiency of the principal component analysis biomarker and the optimal limit of each variable, an ROC analysis method such as AUC (bottom area of ROC curve) was used. P values less than 0.05 were considered statistically significant.
<Result>
First, the surface characteristics of the two substrates used in the present invention, namely, a 50 nm Au-coated anodized aluminum oxide nanodot array substrate and a 2.5 nm Ti- and 50 nm Au-coated Au.0500.ALSI substrate, were observed using AFM, and the results are shown in
For the surface characteristics analysis, NANOS N8 NEOS (Bruker, Germany), which is a tapping mode AFM device, was used, and as a result of analyzing the surface profile of the two types of the SERS substrates used, it was confirmed that the surface roughness characteristics of the 2.5 nm Ti- and 50 nm Au-coated Au.0500.ALSI substrate were reduced by 10 times compared to that of the 50 nm Au-coated anodized aluminum oxide nano-dot array substrate.
SERS activity of the above-mentioned two types of the substrates was observed using a balanced salt solution (BSS) used for eye washing in clinical practice. As a result, as shown in
Although the two types of the SERS substrates exhibited a similar spectral pattern, two-fold stronger intensity was exhibited in an AAO nanodot array substrate. Overall, the AAO nanodot array substrate exhibited more excellent nanostructure and DCD-SERS activity than the commercially available Au.0500.ALSI substrate.
In order to reduce deviations between data by collecting data in various conditions, a pre-processing treatment was performed on the DCD-SERS spectrum. First, representative DCD-SERS spectra of tear samples collected from non-infected persons and adenoviral conjunctivitis-confirmed patients are shown in
As a negative control group, the DCD-SERS spectrum measured using BSS is shown in
In all experiments, 2 μL of tear was used, and the total drying time was 20 minutes, from which dried tear samples having a diameter of approximately 4 mm were obtained. In order to obtain more reliable DCD-SERS spectra for hardware implementation, DCD-SERS spectra were measured and compared according to the different zones of dried tear. As shown in
Furthermore, the characteristics of the DCD-SERS spectrum depending on the amount of tear used from the respective zones were determined, and the results are shown in
As shown in
When
The DCD-SERS spectra measured at 10 different points in the same zone of the same sample were superimposed and shown in
The performance of the AC biomarker in a logarithmic form on tear samples collected from non-infected persons and adenoviral conjunctivitis patients is shown in Table 2, and clinical trial results (n=100, respectively) are shown in Table 3.
200 DCD-SERS spectra measured from the proposed 4 zones were evaluated. First of all, the AC biomarker in tears of non-infected persons exhibited 100% sensitivity and 97% accuracy, regardless of the zone. In the non-infected group, a false positive spectrum was not observed in the C zone, but there were some false positive spectra in the other zones. Meanwhile, adenoviral conjunctivitis patients showed 100% specificity in all zones without false positives, and in the C zone and R zone, accuracies of 100% and 60% were observed, respectively. The error rate in the T zone was half the error rate for the R zone. The AC biomarkers showed a high accuracy of 99% in the C and M zones, and approximately 70% accuracy in the T and R zones.
Further, the AC biomarker depending on the severity of adenoviral conjunctivitis was evaluated. The AC biomarker performance is shown in Table 4 in a logarithmic form according to the severity of adenoviral conjunctivitis, and the clinical test results, which were separated according to the severity, are shown in Table 5. Specifically, whereas the accuracy for mild adenoviral conjunctivitis in the R zone was 27%, in the case of severe adenoviral conjunctivitis, it was 86%, and in the other zones, the accuracy was more than approximately 80%, and in particular, in the C zone, the accuracy was 100%. These results indicate that the zone excluding the outermost ring zone (R zone) is a good Raman spectrum screening region for diagnosing adenoviral infection, and the AC marker in these zones is an excellent parameter for early diagnosis.
Furthermore, three PC loading profiles (PC1, PC2, and PC3) were extracted from information by the tear of the non-infected persons, the tear from the patients with adenoviral conjunctivitis, and the differences of two. This was performed in three DCD-SERS spectral vectors [1242 cm−1, 1342 cm−1], [1242 cm−1, 1448 cm−1], and [1342 cm−1, 1448 cm−1] in the four zones, and the results are shown in
That is, the PCA biomarker showed AUC values of 0.9453 in the C zone and 0.8182 in the R zone, and all of the PCA biomarkers exhibited a high sensitivity of 93% or more and a detection ability of 98% for the non-infected tear samples in the R zone (Table 7). The specificity of the PCA biomarker was 95% in the C zone, 91% in the M zone, 86% in the T zone, and 76% in the R zone. These results indicate that the measurement in the C zone rather than the R zone has an excellent diagnostic ability for adenoviral conjunctivitis. The loading profiles of PC1 and PC2 in the DCD-SERS spectrum accounted for 98% of the total. The passively set linear separating lines (dashed lines) in
The DCD-SERS spectrum measured at wavelengths in a range of 1200 cm−1 to 1500 cm−1 for the C and R zones and the respective 10 Gaussian sub-peaks extracted therefrom are shown in
From low wave numbers, four Gaussian functions of the second Gaussian function (the amide β-sheet at 1242 cm−1), the third function (the amide III α-helix at 1275 cm−1), the fifth function (C—H deformation at 1342 cm−1), and the tenth function (C—H deformation at 1448 cm−1) were selected as peaks for protein analysis in the tear samples.
The characteristics of the MGP biomarkers composed of selected Gaussian functions are summarized in Table 9 below. In the case of the tears from non-infected persons, the amide III β-sheet and C—H deformation were increased by 2 times by progressing from the C zone to the R zone, but the opposite pattern was observed in the tears of the adenoviral conjunctivitis patients. These changes resulted in a significant decrease (p<0.001) in the amide III α-helix of the non-infected group, and a significant increase (p<0.01) in the same of the adenoviral conjunctivitis group, but C—H deformation at 1448 cm−1 did not show any significant difference in the two groups.
Finally, each of the Gaussian functions resolved from the multi-Gaussian model clearly showed differences in the samples collected from the non-infected persons and adenoviral conjunctivitis patients, and this indicates that MGP markers determined by the Gaussian segmentation technique can be used to qualitatively and quantitatively monitor the presence of adenoviral infection. Therefore, the method and system for detecting viral infection of the present invention can be used not only for diagnosing ophthalmic diseases caused by viral infection using tear samples, but also for diagnosing viral infection using other body fluid samples such as saliva, sweat, etc.
Number | Date | Country | Kind |
---|---|---|---|
10-2014-0096695 | Jul 2014 | KR | national |
Filing Document | Filing Date | Country | Kind |
---|---|---|---|
PCT/KR2015/004736 | 5/12/2015 | WO | 00 |