Infrared spectroscopy and multivariate analysis applied to the differential diagnosis of leptospirosis and dengue

Infrared spectroscopy and multivariate analysis applied to the differential diagnosis of leptospirosis and dengue | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Infrared spectroscopy and multivariate analysis applied to the differential diagnosis of leptospirosis and dengue Alejandra Zambrano, Jorge Trilleras, Victoria A. Arana, Kássio M. G. Lima, and 7 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4607844/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Dengue and leptospirosis are considered systemic and very dynamic illnesses in which a patient can rapidly progress from mild to severe conditions. Both diseases present very similar acute initial symptoms, a fact that may result in a challenging differential diagnosis at the initial phases. Herein, we present the application of attenuated total reflectance-Fourier transform infrared (ATR-FTIR) spectroscopy combined with multivariate analysis to perform differential diagnosis of leptospirosis and dengue by analysing blood plasma. The spectra of 114 samples from patients in different phases of infection ( n = 43 for leptospirosis and n = 71 for dengue) were analysed by either linear or quadratic discrimination in association with genetic algorithms, successive projection algorithms and principal component analysis for feature selection/extraction. The best model, GA-QDA, achieved outstanding results in terms of maximum (100%) sensitivity, specificity and accuracy for classifying both classes by using only 31 spectral variables. The ANOVA calculations, at a confidence level of 95%, highlighted a set of 10 variables selected by the GA-QDA model (1296 cm -1 , 1612 cm -1 , 1673 cm -1 , 1677 cm -1 , 1678 cm -1 , 1689 cm -1 , 1694 cm -1 , 1711 cm -1 , 1713 cm -1 and 1719 cm -1 ) with significant differences in the absorbance means between the Leptospirosis and Dengue classes. These specific wavenumbers represent the most useful spectral information accounting for the biochemical changes that mark a specific infection. These remarkable results obtained in this pilot study highlight the viability of this methodology to be applied in clinical practice to serve as a simple and accurate test for discriminating between the two illnesses. leptospirosis dengue differential diagnosis ATR-FTIR multivariate analysis Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction Dengue and leptospirosis are considered systemic and very dynamic illnesses in which a patient can rapidly progress from mild to severe conditions. In general, these diseases manifest abruptly with nonspecific acute symptoms, mostly characterized by high-grade fever (39–40°C), myalgia, headaches, chills, loss of appetite, nausea, vomiting and diarrhea. In this initial stage, known as the feverish phase, which usually lasts from 3 to 7 days, the immune response of individuals can be quite diverse because it is often complex to predict how the disease is expected to evolve 1 . Although the symptoms and evolution of both illnesses are considerably similar, the treatment strategies commonly applied for dengue and leptospirosis are essentially different since they are diseases caused by viral and bacterial agents, respectively. The success of treatment is highly dependent on the stage in which the disease is diagnosed, and if not started in a timely manner, severe damage to organs and hemorrhage can occur, dramatically increasing the risk of death of the patient. Early diagnosis has outstanding importance in managing medical intervention and starting treatment; however, unfortunately, in many real cases, the differential diagnosis between dengue and leptospirosis can be challenging given the clinical similarities between the two conditions, which can ultimately lead to inappropriate medical care 2 . Currently, the most commonly applied diagnostic techniques for detecting dengue or leptospirosis are based on laboratory tests since clinical evaluation can easily be prone to misinterpretation 3 . Traditional methods include pathogen isolation, genome detection, and antigen/antibody serologic and immunoassay tests; however, it should be noted that these methods may be subject to loss of specificity compared with other similar pathogens, such as serologic tests for dengue and Zika viruses 4,5 . Additionally, for leptospirosis, these diagnostic tools are highly affected by their sensitivity rates, which can reach values as low as 40% as a consequence of the sample used (mostly blood or urine) as well as in which phase of the disease (days since disease onset) the samples were obtained 6,7 . Currently, there is no gold standard test that can be used to unequivocally diagnose both diseases because their detection is usually based on a combination of available laboratory tests and clinical/epidemiological evaluation. In light of these facts, the development of alternative methods that could contribute to the differential diagnosis of these neglected illnesses is highly important, especially for endemic regions, such as the Caribbean and Latin America, which normally face outbreaks and high mortality rates. Biospectroscopy can be understood as the use of a spectroscopic technique to analyse biological samples. This spectroscopic approach has gained great attention in the medical field as a potential tool for the detection and diagnosis of diseases 8–11 . In this context, infrared spectroscopy plays an important role as one of the most applied techniques, with a great number of reported successful studies found in the literature in recent decades. Attenuated total reflection-Fourier transform infrared (ATR-FTIR) spectroscopy offers a simple and fast methodology for analysing a wide range of biological samples since the specimens (such as biofluids, cells, and tissues) can be placed directly on the ATR crystal, and no prior sample treatment is needed. The infrared spectra are the result of the molecular vibrations of the constituents of the sample, which can interact with the infrared beam and absorb parts of it. The range between 1800 and 900 cm − 1 , called the biofingerprint region, provides very rich information about the composition of biological samples, comprehending important vibrational bands related to functional groups of lipids (~ 1750 cm − 1 ), proteins amide I, II and III (~ 1650 cm − 1 , ~ 1550 cm − 1 and ~ 1260 cm − 1 , respectively), carbohydrates (~ 1150 cm − 1 ) and RNA/DNA (~ 1080 cm − 1 and 1225 cm − 1 ), for example 12 . With the understanding that every disease promotes specific immune responses in individuals at biochemical levels, infrared spectra can be seen as powerful sources of physiopathological information to be analysed, aiming to detect and diagnose diseases based on their chemical alterations reflected in biological samples. Despite being very informative, the infrared spectra of biological samples can also be complex in terms of interpretation since a large number of different molecules can contribute to the signal obtained, which may be a drawback of its application to real case scenarios. To address this complexity, multivariate analysis is carried out in combination with ATR-FTIR spectroscopy in most cases and provides a wide variety of chemometric methods to be applied for different purposes. With the aim of detecting and diagnosing diseases, multivariate classification methods such as linear discriminant analysis (LDA), partial least squares discriminant analysis (PLS-DA) and support vector machines (SVMs) have proven the potential of coupling spectroscopic and chemometric methods to screen and diagnose illnesses in many applications 13–16 . Santos and colleagues conducted a pioneering study to differentiate healthy individuals from those suffering from dengue, Zika or chikungunya infections through the analysis of blood plasma samples by ATR-FTIR spectroscopy. Combined with algorithms for linear discriminant analysis, they could classify samples with remarkable sensitivity and specificity of 100% for healthy, dengue and chikungunya classes and values close to 90% for zika samples 12 . A novel approach was proposed by Naseer and colleagues for dengue diagnosis, in which freeze-dried blood serum samples were subjected to ATR-FTIR analysis as a diagnostic tool to identify biochemical markers of dengue infection and to distinguish between healthy and diseased samples. The sensitivity and specificity values obtained for the classification were 89% and 95%, respectively 17 . By expanding their methodology for analysing freeze-dried blood sera, Ali and coworkers investigated the use of ATR-FTIR spectroscopy for the differential diagnosis of dengue and hepatitis C infections in human blood. With the aid of multivariate analysis, the authors reported a classification accuracy of 99.2%, providing a viable alternative method for the differential diagnosis of viral diseases 10 . These recent findings demonstrate the suitability of ATR-FTIR spectroscopy coupled with multivariate analysis for analysing biological samples, in special blood plasma/sera, not only for identifying the presence or absence of diseases but also for the development of alternative methods for differential diagnosis. Therefore, the main purpose of this pilot study was to develop a simple, fast and low-cost methodology based on ATR-FTIR spectroscopy and multivariate classification for the accurate differential diagnosis of leptospirosis and dengue in humans via blood plasma analysis. Considering that both diseases present nonspecific acute symptoms that may lead to misinterpretation and incorrect diagnosis, there is still a need for alternative methods that can improve diagnosis and, consequently, the correct medical management of patients at initial phases, which contributes to decreasing both the mortality rates and the risks of outbreaks, especially in endemic regions. Methods Collection and preparation of specimens This study involved patients living in Colombia who were admitted to different hospitals for medical assistance and treatment for symptoms of infection. All samples were collected and processed by the Laboratório de Salud Pública of the Departamento del Atlántico. All experiments were approved by the medical ethics committee at the Universidad del Atlántico (Colombia), with an endorsement letter emitted in August 2023. Informed consent was obtained from each human participant and all methods were performed in accordance with the relevant guidelines and regulations (Declaration of Helsinki). Blood plasma samples from 114 patients in different phases of infection (n = 43 for leptospirosis and n = 71 for dengue), between men and women aged five to sixty years, were collected by venipuncture and transferred to tubes containing EDTA as an anticoagulant. Within 2 hours after collection, blood plasma samples were separated by density gradient, and aliquots were placed into cryogenic tubes and stored at -80°C until analysis. Leptospirosis was diagnosed by the reference serological test MAT (microscopic agglutination test) and ELISA for IgM. Dengue diagnosis was carried out by ELISA tests for IgM and NS1 (dengue antigen protein). According to the inclusion criteria, only samples collected within 18 months after blood collection were considered in this study. ATR-FTIR spectroscopy ATR-FTIR spectral measurements were performed using a JASCO 4700 FTIR spectrometer (JASCO Corporation, Tokyo, Japan) with an ATR attachment containing a diamond crystal internal reflective element at a 45° incidence angle of the IR beam. The instrument was set up to perform a total of 16 scans with 4 cm -1 spectral resolution on both the background and sample. The ATR crystal was cleaned with 98% isopropanol, and before each sample spectral acquisition, a background was collected. An aliquot of 20 µL of each room temperature blood plasma sample was placed onto aluminum foil as the substrate for spectral acquisition, according to Cui and colleagues 18 . Each sample was analysed in triplicate, resulting in a total of 342 spectra (n = 129 for leptospirosis and n = 213 for dengue). Data analysis and chemometric methods All computational analyses, including importing, preprocessing and construction of multivariate classification models, were performed within the MATLAB® 2022 environment (MathWorks Inc., Natick, MA, USA) by using PLS Toolbox version 9.2 (Eigenvector Research Inc., USA) and laboratory-made routines. Initially, every plasma sample was analysed as its mean spectra, which were calculated using the three spectral replicates obtained for the sample, resulting in 43 mean spectra for leptospirosis samples and 71 mean spectra for dengue samples. Additionally, the ATR-FTIR mean spectra were preprocessed by selecting the fingerprint region between 1900 and 1000 cm − 1 , resulting in a 114 x 935 matrix that was baseline corrected (automatic weighted least squares) and normalized to the amide I peak (i.e., ~ 1650 cm − 1 ). For modelling, the samples were divided into training (~ 70%, n = 77), validation (~ 15%, n = 18) and prediction (~ 15%, n = 18) sets by applying the classical Kennard-Stone algorithm to the ATR-FTIR mean spectra 19 . One sample of the leptospirosis class was excluded and not modelled based on its outlier spectral behaviour. The training samples were used for model building and optimization (feature selection by the SPA and GA algorithms), the validation set was used to evaluate the performance of the models, and the prediction set was applied to test the optimized models in a real case situation when dealing with unknown samples. The discrimination models PCA-LDA, PCA-QDA, SPA-LDA, SPA-QDA, GA-LDA and GA-QDA were built by using laboratory-made routines for MATLAB software. A detailed description of the algorithms has already been provided by our research group and can be found elsewhere 20,21 . The GA-LDA/QDA calculations were performed for 40 generations with 80 chromosomes each. The one-point crossover and mutation probabilities were set to 60% and 10%, respectively. Moreover, the algorithm was repeated three times, starting from different random initial populations. The best solution (in terms of the fitness value) resulting from the three GA repetitions was employed. Calculations of figures of merit, including sensitivity (probability that a test result will be positive when disease is present), specificity (probability that a test result will be negative when disease is not present), accuracy (number of samples correctly classified considering true and false negatives) and both the area under the ROC curve (AUC) and F Score, for indicating model performance, were obtained for this study as important quality measures to evaluate the discriminant ability of the models when applied to unknown/test samples. These parameters have a maximum value of 1 (100%) and a minimum of 0 and can be obtained by using the following equations 22,23 : where FN is defined as a false negative and FP is defined as a false positive. TP and TN are defined as true positive and true negative, respectively. Herein, for all calculations, leptospirosis samples were considered the positive class (“disease group”), and dengue samples were considered the negative class (“control group”). Results In this study, a total of 114 blood plasma samples from patients suffering from acute symptoms of dengue (n = 71) or leptospirosis (n = 43) were analysed by infrared spectroscopy using the ATR mode, directly with no need for sample pretreatment. The raw spectra of all the samples are shown in Fig. 1 -A in the fingerprint region between 1900 and 1000 cm − 1 , where important biomolecules present significant absorption bands. Previously, for the construction of the discrimination models, spectra were preprocessed with baseline correction and normalization to the amide I peak to remove physical interferences and highlight chemical signals between the classes, as shown in Fig. 1 -B. The mean preprocessed spectra of both classes, Dengue and Leptospirosis, are presented in Fig. 1 -C, which shows strong similarity in their absorption bands. [Insert Fig. 1 A, B and C here] Figure 2 shows the score plot for PCA of the dataset, with 95.67% of the explained variance accounting for the first two principal components. As the first attempt to observe possible clusters within samples, PCA analysis revealed no clear separation between samples belonging to the dengue or leptospirosis classes. On the other hand, there is an important observation in this graphic related to the presence of an outlier sample of the leptospirosis class that lies very far from the confidence limits. This sample was removed from the dataset before the construction of the discrimination models, resulting in a matrix of size 113 x 935. [Insert Fig. 2 here] For modelling, samples were split into three sets for calibration, validation and prediction according to the calculations of the Kennard-Stone algorithm. Linear and quadratic discrimination models were constructed by using either the scores of principal components (PCA-LDA and PCA-QDA) or a subset of selected spectral variables obtained by SPA or GA algorithms (SPA-LDA, SPA-QDA, GA-LDA, GA-QDA). Table 1 represents the confusion matrix for each model, considering the samples of the prediction set. Table 1 – Confusion matrices for the discrimination models for dengue and leptospirosis using the prediction set. Model Actual Class Dengue Leptospirosis PCA-LDA ( 3 ) * Dengue 7 1 Leptospirosis 4 6 SPA-LDA ( 2 ) Dengue 8 5 Leptospirosis 3 2 GA-LDA (33) Dengue 9 3 Leptospirosis 2 4 PCA-QDA ( 3 ) Dengue 7 1 Leptospirosis 4 6 SPA-QDA ( 2 ) Dengue 7 3 Leptospirosis 4 4 GA-QDA (31) Dengue Leptospirosis 11 0 0 7 * Number of principal components or selected variables is shown in parentheses. [Insert Table 1 here] For the LDA models, GA-LDA and SPA-LDA provided poor classification for the leptospirosis class, while PCA-LDA, which uses the scores of three principal components, misclassified only one sample in that class. On the other hand, for discriminating dengue patients, SPA and GA-LDA outperformed PCA-LDA. Quadratic discrimination did not enhance the classification performance of PCA-QDA for either of the two classes; however, it provided better results for SPA-QDA for leptospirosis samples. In the case of GA-QDA, an outstanding improvement in classification was observed for both classes, with the model reaching 100% correct classification for all prediction samples, using only 31 spectral variables. To provide a better evaluation of the model’s performance, some figures of merit, including sensitivity, specificity, accuracy, F-Score and AUC calculations, are depicted in Table 2 . Table 2 – Figures of merit for the discrimination models for dengue and leptospirosis using the prediction set. Model Sensitivity (%) Specificity (%) Accuracy F Score AUC PCA-LDA 85.7 63.6 72.2 0.730 0.746 SPA-LDA 28.5 72.7 55.5 0.409 0.506 GA-LDA 57.1 81.8 72.2 0.672 0.694 PCA-QDA 85.7 63.6 72.2 0.730 0.746 SPA-QDA 57.1 63.6 61.1 0.601 0.603 GA-QDA 100 100 100 1.00 1.00 [Insert Table 2 here] Considering leptospirosis as the positive class, the sensitivity of the models (meaning their ability to correctly identify patients positive for leptospirosis infection) for the prediction set ranged from 28.5–100%. The specificity (i.e., the ability of the models to identify patients positive for dengue infection) varied from 63.6–100%. The accuracy, which reflects the performance of the discrimination models, ranged from 55.5–100%. The best model for linear discrimination was PCA-LDA, with an accuracy of 72.2% and F- score and AUC of 0.730 and 0.746, respectively. When using QDA, the GA-QDA model presented remarkable quality parameters for classification, reaching maximum rates for all the metrics calculated herein. The discriminant function showing the separation between leptospirosis and dengue samples is depicted in Fig. 3 -A for the best model, GA-QDA. The selected subset of 31 out of 935 variables is shown in Fig. 3 -B, highlighting the main spectral regions used for correct discrimination. The area under the ROC curve is plotted in Fig. 3 -C for both the GA-LDA and GA-QDA models to demonstrate the improvement in the performance of the GA model when the quadratic discrimination function was applied. [Insert Fig. 3 -A, B and C here] Analysis of variance was applied to the 31 variables chosen by the GA-QDA model to verify significant differences between the means of the absorption signals of the dengue and leptospirosis samples. At a confidence level of 95%, a set of 10 out of 31 variables presented significant differences (P < 0.05) in the ANOVA calculations, as shown in Fig. 4 . [Insert Fig. 4 here] According to the box plots, the mean absorbance values of the dengue samples were greater than those of the leptospirosis samples for all the variables except for the wavenumber at 1612 cm − 1 . Additionally, more outliers were identified for the Dengue class, while for Leptospirosis, a more evident spread was observed, with larger boxes as well as whiskers in a larger range, excluding the wavenumber at 1296 cm − 1 . Discussion Leptospirosis and dengue are both infectious diseases that present very similar acute initial symptoms and usually include high-grade fever, strong headache, nausea, chills and myalgia. Such similarity in the acute phase may result in additional risks for patients, considering that the differential diagnosis of both illnesses may be challenging and that medical care should be specific for each disease since they are viral (dengue) and bacterial (leptospirosis) infections. A misleading diagnosis not only worsens the prognosis since these infections can evolve rapidly for severe conditions but also impairs control and mitigation in endemic areas. The use of spectroscopic techniques to analyse biological samples, known as biospectroscopy, has gained great attention in recent decades since it allows the investigation of diseases based on biochemical changes that occur as a consequence of pathological events 24 . Vibrational spectroscopies have been successfully applied in many studies in medical chemistry 25 since they provide alternative methods for screening, identification and diagnosis of diseases through simple, cost-effective and accurate methodologies. Herein, we used different chemometric methods to analyse ATR-FTIR data for the differential diagnosis of leptospirosis and dengue in patients from Colombia, an endemic country for both diseases. With the exploratory analysis of PCA, no clear separation was observed in the samples; however, an outlier could be visually identified. This unique sample refers to a nonpathogenic strain of Leptospira known as saprophytic, as confirmed by the reference MAT test. This finding demonstrates the usefulness of PCA for identifying patterns and improving comprehension of the dataset. Classification models based on linear or quadratic discriminant analysis were constructed to allow the differential diagnosis of leptospirosis from dengue. With LDA, the best accuracy obtained was 72.2% by the PCA-LDA and GA-LDA models, representing an unsatisfactory ability to differentiate the diseases. Quadratic discrimination by the GA-QDA model considerably improved the classification results, achieving 100% accuracy, sensitivity and specificity, which allows the differential diagnosis of leptospirosis and dengue. These results highlight that when analysing imbalanced datasets (both classes do not have the same number of samples), quadratic discrimination should be the best suited tool. The 31 main variables responsible for discriminating the two classes, selected by GA, were compared by using analysis of variance. The ANOVA results, at a confidence level of 95%, showed that only 10 variables (1296 cm -1 , 1612 cm -1 , 1673 cm -1 , 1677 cm -1 , 1678 cm -1 , 1689 cm -1 , 1694 cm -1 , 1711 cm -1 , 1713 cm -1 and 1719 cm -1 ) presented significant differences in absorbance between the leptospirosis and dengue classes. These specific wavenumbers represent the most useful spectral information accounting for the biochemical changes that mark a specific infection. Some molecular bonds can be tentatively assigned to the ten significant variables, such as the band of amide III (~ from 1290 to 1334 cm -1 ), the bands of amide I (C = N associated with β-sheets, random coil and C = O stretch of conjugated ketones) in the region of approximately 1620 to 1690 cm -1 , and the region from 1700 to 1720 related to the C = O stretching vibrations of ketones and aldehydes 26,27 . The differential diagnosis between leptospirosis and dengue by infrared spectroscopy and multivariate analysis is based on the biochemical variations that occur as a consequence of the pathological process. The spectral regions responsible for the discrimination of samples cannot be directly associated with specific molecules, such as biomarkers; however, they represent important information about the classes of biomolecules most affected by these illnesses. Therefore, other techniques, such as mass spectrometry, should be more suitable for the investigation of possible biomarkers, and a larger number of samples should be included in further studies. Nonetheless, the spectroscopic-multivariate method presented herein provides a very useful alternative for overcoming the challenges related to leptospirosis and dengue infections as a simple, cost-effective and accurate methodology for differential diagnosis. Conclusion Herein, we demonstrated the use of ATR-FTIR spectroscopy combined with multivariate analysis for the differential diagnosis of two infectious diseases, leptospirosis and dengue. By quadratic discriminant analysis, the GA-QDA model reached maximum rates of correct classification, with sensitivity, specificity and accuracy values of 100% when analysing test (unknown) samples. The outstanding results obtained in this pilot study highlight the capability of this methodology to be applied in clinical practice to serve as a rapid and accurate test for discriminating between the two illnesses. Declarations Acknowledgements The authors of this work would like to thank the Chemistry Postgraduation Program of the Federal University of Rio Grande do Norte. Kássio M.G. Lima thanks the National Council for Scientific and Technological Development grant (Ref. 305562/2020-7). Data Availability The datasets used and/or analysed during the current study available from the corresponding author on reasonable request. References Rodríguez-Salazar, C. A. et al. Manifestaciones clínicas y hallazgos de laboratorio de una serie de casos febriles agudos con diagnóstico presuntivo de infección por el virus dengue. Quindío (Colombia). Infectio 20 , 84–92 (2016). Hartskeerl, R. A., Collares-Pereira, M. & Ellis, W. A. Emergence, control and re-emerging leptospirosis: Dynamics of infection in the changing world. Clinical Microbiology and Infection vol. 17 494–501 Preprint at https://doi.org/10.1111/j.1469-0691.2011.03474.x (2011). Clemen G, Angel J, Montes C, Jr, T. & Osorio L. Contribución de La Prueba Rápida NS1 e IgM al Diagnóstico de Dengue En Colombia En El Periodo Pre-Zika . Peeling, R. W. et al. Evaluation of diagnostic tests: Dengue. Nat Rev Microbiol 8 , S30–S38 (2010). Gárcia, J. D., Agûero, J., Parra, JA. & Santos, MF. Enfermedades infecciosas. Concepto. Clasificación. Aspectos generales y específicos de las infecciones. Criterios de sospecha de enfermedad infecciosa. Pruebas diagnósticas complementarias. Criterios de indicación. Medicina - Programa de formaciòn Medica Continuada acreditado 10 , 3251–3264 (2010). Pinto, G. V. et al. Current methods for the diagnosis of leptospirosis: Issues and challenges. Journal of Microbiological Methods vol. 195 Preprint at https://doi.org/10.1016/j.mimet.2022.106438 (2022). Suwannin, P. et al. Enhancing leptospirosis control with nanosensing technology: A critical analysis. Comparative Immunology, Microbiology and Infectious Diseases vol. 104 Preprint at https://doi.org/10.1016/j.cimid.2023.102092 (2024). Santos, M. C. D. et al. ATR-FTIR spectroscopy with chemometric algorithms of multivariate classification in the discrimination between healthy: Vs. dengue vs. chikungunya vs. zika clinical samples. Analytical Methods 10 , 1280–1285 (2018). Naseer, K., Ali, S. & Qazi, J. ATR-FTIR spectroscopy based differentiation of typhoid and dengue fever in infected human sera. Infrared Phys Technol 114 , (2021). Ali, Salmann., Naseer, Khulla., Hussain, Irshad. & Qazi, Javaria. ATR-FTIR spectroscopy-based differentiation of hepatitis C and dengue infection in human freeze-dried sera. Infrared Phys Technol 118 , 1350–4495 (2021). Morais, C. L. M., Lima, K. M. G., Singh, M. & Martin, F. L. Tutorial: multivariate classification for vibrational spectroscopy in biological samples. Nature Protocols vol. 15 2143–2162 Preprint at https://doi.org/10.1038/s41596-020-0322-8 (2020). Santos, M. C. D. et al. ATR-FTIR spectroscopy with chemometric algorithms of multivariate classification in the discrimination between healthy: Vs. dengue vs. chikungunya vs. zika clinical samples. Analytical Methods 10 , 1280–1285 (2018). Kawakami, R. et al. Aspects of the Successive Projections Algorithm for Variable Selection in Multivariate Calibration Applied to Plasma Emission Spectrometry . Analytica Chimica Acta vol. 443 (2001). Silva, H. K. T. de A. et al. Detection of terbufos in cases of intoxication by means of entomotoxicological analysis using ATR-FTIR spectroscopy combined with chemometrics. Acta Trop 238 , (2023). Dixon, S. J. & Brereton, R. G. Comparison of performance of five common classifiers represented as boundary methods: Euclidean Distance to Centroids, Linear Discriminant Analysis, Quadratic Discriminant Analysis, Learning Vector Quantization and Support Vector Machines, as dependent on data structure. Chemometrics and Intelligent Laboratory Systems 95 , 1–17 (2009). Morais, C. L. M. & Lima, K. M. G. Principal component analysis with linear and quadratic discriminant analysis for identification of cancer samples based on mass spectrometry. J Braz Chem Soc 29 , 472–481 (2018). Salmann, A., Khulla, N., Irshad, H. & Javaria, Q. ATR-FTIR spectroscopy-based differentiation of hepatitis C and dengue infection in human freeze-dried sera. Infrared Physics and Technology 1–4 (2021). Cui, L., Butler, H. J., Martin-Hirsch, P. L. & Martin, F. L. Aluminium foil as a potential substrate for ATR-FTIR, transflection FTIR or Raman spectrochemical analysis of biological specimens. Analytical Methods 8 , 481–487 (2016). Kennard, R. W. & Stone, L. A. Computer Aided Design of Experiments . vol. 11 http://www.jstor.org/stable/1266770 . (1969). Morais, C. L. M., Lima, K. M. G., Singh, M. & Martin, F. L. Tutorial: multivariate classification for vibrational spectroscopy in biological samples. Nature Protocols vol. 15 2143–2162 Preprint at https://doi.org/10.1038/s41596-020-0322-8 (2020). Neves, A. C. O. et al. ATR-FTIR and multivariate analysis as a screening tool for cervical cancer in women from northeast Brazil: A biospectroscopic approach. RSC Adv 6 , 99648–99655 (2016). Morais, C. L. M., Lima, K. M. G., Singh, M. & Martin, F. L. Tutorial: multivariate classification for vibrational spectroscopy in biological samples. Nature Protocols vol. 15 2143–2162 Preprint at https://doi.org/10.1038/s41596-020-0322-8 (2020). López, V., Fernández, A., García, S., Palade, V. & Herrera, F. An insight into classification with imbalanced data: Empirical results and current trends on using data intrinsic characteristics. Inf Sci (N Y) 250 , 113–141 (2013). da Silva, T. G. et al. Spectrochemical analysis of blood combined with chemometric techniques for detecting osteosarcopenia. Sci Rep 13 , (2023). Alkhuder, K. Fourier transform infrared spectroscopy: A universal optical sensing technique with auspicious application prospects in the diagnosis and management of autoimmune diseases. Photodiagnosis and Photodynamic Therapy vol. 42 Preprint at https://doi.org/10.1016/j.pdpdt.2023.103606 (2023). Walsh, M. J. et al. IR microspectroscopy: potential applications in cervical cancer screening. Cancer Letters vol. 246 1–11 Preprint at https://doi.org/10.1016/j.canlet.2006.03.019 (2007). Purandare, N. C. et al. Infrared spectroscopy with multivariate analysis segregates low-grade cervical cytology based on likelihood to regress, remain static or progress. Analytical Methods 6 , 4576–4584 (2014). Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4607844","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":326050077,"identity":"ba8b5946-c157-4ee4-abca-21989d248ff8","order_by":0,"name":"Alejandra Zambrano","email":"data:image/png;base64,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","orcid":"","institution":"Universidad del Atlántico","correspondingAuthor":true,"prefix":"","firstName":"Alejandra","middleName":"","lastName":"Zambrano","suffix":""},{"id":326050078,"identity":"f270fda8-21bd-4d62-b3c7-3a8681c0b230","order_by":1,"name":"Jorge Trilleras","email":"","orcid":"","institution":"Universidad del Atlántico","correspondingAuthor":false,"prefix":"","firstName":"Jorge","middleName":"","lastName":"Trilleras","suffix":""},{"id":326050079,"identity":"b9ed532b-e1eb-4c13-b8ab-ba7241edddff","order_by":2,"name":"Victoria A. Arana","email":"","orcid":"","institution":"Universidad del Atlántico","correspondingAuthor":false,"prefix":"","firstName":"Victoria","middleName":"A.","lastName":"Arana","suffix":""},{"id":326050080,"identity":"80d2988b-2c6a-4fa1-8c2f-3eb5a8e6a18c","order_by":3,"name":"Kássio M. G. Lima","email":"","orcid":"","institution":"Federal University of Rio Grande do Norte","correspondingAuthor":false,"prefix":"","firstName":"Kássio","middleName":"M. G.","lastName":"Lima","suffix":""},{"id":326050081,"identity":"074bac2b-5cca-4c64-8fa1-2c04bb08e9c8","order_by":4,"name":"Ana C. O. Neves","email":"","orcid":"","institution":"Federal University of Rio Grande do Norte","correspondingAuthor":false,"prefix":"","firstName":"Ana","middleName":"C. O.","lastName":"Neves","suffix":""},{"id":326050082,"identity":"410868f7-a0b6-4d7c-9299-a270129bd6cd","order_by":5,"name":"Camilo L. M. Morais","email":"","orcid":"","institution":"Federal University of Rio Grande do Norte","correspondingAuthor":false,"prefix":"","firstName":"Camilo","middleName":"L. M.","lastName":"Morais","suffix":""},{"id":326050083,"identity":"b8fcc824-141a-4d71-9248-b9969d2fd6d9","order_by":6,"name":"Claudia Romero","email":"","orcid":"","institution":"Universidad del Norte","correspondingAuthor":false,"prefix":"","firstName":"Claudia","middleName":"","lastName":"Romero","suffix":""},{"id":326050085,"identity":"3d09438f-5bb4-4b32-aa44-68044be5a13a","order_by":7,"name":"Andrew K. I. Falconar","email":"","orcid":"","institution":"Universidad del Norte","correspondingAuthor":false,"prefix":"","firstName":"Andrew","middleName":"K. I.","lastName":"Falconar","suffix":""},{"id":326050088,"identity":"f2a13bf8-2efa-445a-a17f-d18058262f89","order_by":8,"name":"Boanegre S. Muñoz","email":"","orcid":"","institution":"Universidad del Atlántico","correspondingAuthor":false,"prefix":"","firstName":"Boanegre","middleName":"S.","lastName":"Muñoz","suffix":""},{"id":326050089,"identity":"d04bc0ca-9ae1-4544-aa20-3fec1a3dca06","order_by":9,"name":"Roberto García","email":"","orcid":"","institution":"Universidad del Atlántico","correspondingAuthor":false,"prefix":"","firstName":"Roberto","middleName":"","lastName":"García","suffix":""},{"id":326050090,"identity":"cd8cf7d6-fde4-40c9-8fa0-3ac6073b0e86","order_by":10,"name":"Carlos Carmona","email":"","orcid":"","institution":"Laboratorio de Salud Pública del Departamento del Atlántico","correspondingAuthor":false,"prefix":"","firstName":"Carlos","middleName":"","lastName":"Carmona","suffix":""}],"badges":[],"createdAt":"2024-06-19 20:09:42","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4607844/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4607844/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":60619997,"identity":"f5c0ef1e-df5b-4aaa-a433-613722e3b943","added_by":"auto","created_at":"2024-07-18 20:47:51","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":192347,"visible":true,"origin":"","legend":"\u003cp\u003eRaw (A), preprocessed(B) and mean spectra (C) of leptospirosis and dengue samples.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-4607844/v1/7b01f4e5072af1d3f19e9fb0.png"},{"id":60619996,"identity":"9c1e1dd1-3ffe-4c23-98b8-d92c17df3d3f","added_by":"auto","created_at":"2024-07-18 20:47:51","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":178196,"visible":true,"origin":"","legend":"\u003cp\u003ePCA scoreplot for the preprocesseddataset. Green squares and red diamonds indicate leptospirosis and denguesamples, respectively.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-4607844/v1/466864d6105f2e8ca82ac805.png"},{"id":60619998,"identity":"38211013-8845-4e74-b08b-59fb3388029f","added_by":"auto","created_at":"2024-07-18 20:47:52","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":178069,"visible":true,"origin":"","legend":"\u003cp\u003eDiscrimination plot (A) and selected variables (B)\u003cstrong\u003e \u003c/strong\u003efor the GA-QDA model; ROC curves for both the GA-LDA and GA-QDA models (C). The blue circles indicate dengue,and the red circles indicateleptospirosis.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-4607844/v1/5d8eecdc566d0ecf28f31396.png"},{"id":60619999,"identity":"7f1642e1-503d-4bb1-8d9e-09df0667a9bf","added_by":"auto","created_at":"2024-07-18 20:47:52","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":223902,"visible":true,"origin":"","legend":"\u003cp\u003eANOVA calculations for the selected variables by the GA-QDA model.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-4607844/v1/546889c58cd010590370fe90.png"},{"id":61573164,"identity":"cab38a0d-b5a8-47c0-8f28-6c9ef3dc8eeb","added_by":"auto","created_at":"2024-08-01 11:32:46","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1238471,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4607844/v1/f0df78df-494d-4519-aaad-2a883750ebed.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Infrared spectroscopy and multivariate analysis applied to the differential diagnosis of leptospirosis and dengue","fulltext":[{"header":"Introduction","content":"\u003cp\u003eDengue and leptospirosis are considered systemic and very dynamic illnesses in which a patient can rapidly progress from mild to severe conditions. In general, these diseases manifest abruptly with nonspecific acute symptoms, mostly characterized by high-grade fever (39\u0026ndash;40\u0026deg;C), myalgia, headaches, chills, loss of appetite, nausea, vomiting and diarrhea. In this initial stage, known as the feverish phase, which usually lasts from 3 to 7 days, the immune response of individuals can be quite diverse because it is often complex to predict how the disease is expected to evolve \u003csup\u003e1\u003c/sup\u003e. Although the symptoms and evolution of both illnesses are considerably similar, the treatment strategies commonly applied for dengue and leptospirosis are essentially different since they are diseases caused by viral and bacterial agents, respectively. The success of treatment is highly dependent on the stage in which the disease is diagnosed, and if not started in a timely manner, severe damage to organs and hemorrhage can occur, dramatically increasing the risk of death of the patient. Early diagnosis has outstanding importance in managing medical intervention and starting treatment; however, unfortunately, in many real cases, the differential diagnosis between dengue and leptospirosis can be challenging given the clinical similarities between the two conditions, which can ultimately lead to inappropriate medical care \u003csup\u003e2\u003c/sup\u003e. Currently, the most commonly applied diagnostic techniques for detecting dengue or leptospirosis are based on laboratory tests since clinical evaluation can easily be prone to misinterpretation \u003csup\u003e3\u003c/sup\u003e. Traditional methods include pathogen isolation, genome detection, and antigen/antibody serologic and immunoassay tests; however, it should be noted that these methods may be subject to loss of specificity compared with other similar pathogens, such as serologic tests for dengue and Zika viruses \u003csup\u003e4,5\u003c/sup\u003e. Additionally, for leptospirosis, these diagnostic tools are highly affected by their sensitivity rates, which can reach values as low as 40% as a consequence of the sample used (mostly blood or urine) as well as in which phase of the disease (days since disease onset) the samples were obtained \u003csup\u003e6,7\u003c/sup\u003e. Currently, there is no gold standard test that can be used to unequivocally diagnose both diseases because their detection is usually based on a combination of available laboratory tests and clinical/epidemiological evaluation. In light of these facts, the development of alternative methods that could contribute to the differential diagnosis of these neglected illnesses is highly important, especially for endemic regions, such as the Caribbean and Latin America, which normally face outbreaks and high mortality rates.\u003c/p\u003e \u003cp\u003eBiospectroscopy can be understood as the use of a spectroscopic technique to analyse biological samples. This spectroscopic approach has gained great attention in the medical field as a potential tool for the detection and diagnosis of diseases \u003csup\u003e8\u0026ndash;11\u003c/sup\u003e. In this context, infrared spectroscopy plays an important role as one of the most applied techniques, with a great number of reported successful studies found in the literature in recent decades. Attenuated total reflection-Fourier transform infrared (ATR-FTIR) spectroscopy offers a simple and fast methodology for analysing a wide range of biological samples since the specimens (such as biofluids, cells, and tissues) can be placed directly on the ATR crystal, and no prior sample treatment is needed. The infrared spectra are the result of the molecular vibrations of the constituents of the sample, which can interact with the infrared beam and absorb parts of it. The range between 1800 and 900 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, called the biofingerprint region, provides very rich information about the composition of biological samples, comprehending important vibrational bands related to functional groups of lipids (~\u0026thinsp;1750 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e), proteins amide I, II and III (~\u0026thinsp;1650 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, ~\u0026thinsp;1550 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e and ~\u0026thinsp;1260 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, respectively), carbohydrates (~\u0026thinsp;1150 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) and RNA/DNA (~\u0026thinsp;1080 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e and 1225 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e), for example \u003csup\u003e12\u003c/sup\u003e. With the understanding that every disease promotes specific immune responses in individuals at biochemical levels, infrared spectra can be seen as powerful sources of physiopathological information to be analysed, aiming to detect and diagnose diseases based on their chemical alterations reflected in biological samples. Despite being very informative, the infrared spectra of biological samples can also be complex in terms of interpretation since a large number of different molecules can contribute to the signal obtained, which may be a drawback of its application to real case scenarios. To address this complexity, multivariate analysis is carried out in combination with ATR-FTIR spectroscopy in most cases and provides a wide variety of chemometric methods to be applied for different purposes. With the aim of detecting and diagnosing diseases, multivariate classification methods such as linear discriminant analysis (LDA), partial least squares discriminant analysis (PLS-DA) and support vector machines (SVMs) have proven the potential of coupling spectroscopic and chemometric methods to screen and diagnose illnesses in many applications \u003csup\u003e13\u0026ndash;16\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eSantos and colleagues conducted a pioneering study to differentiate healthy individuals from those suffering from dengue, Zika or chikungunya infections through the analysis of blood plasma samples by ATR-FTIR spectroscopy. Combined with algorithms for linear discriminant analysis, they could classify samples with remarkable sensitivity and specificity of 100% for healthy, dengue and chikungunya classes and values close to 90% for zika samples \u003csup\u003e12\u003c/sup\u003e. A novel approach was proposed by Naseer and colleagues for dengue diagnosis, in which freeze-dried blood serum samples were subjected to ATR-FTIR analysis as a diagnostic tool to identify biochemical markers of dengue infection and to distinguish between healthy and diseased samples. The sensitivity and specificity values obtained for the classification were 89% and 95%, respectively \u003csup\u003e17\u003c/sup\u003e. By expanding their methodology for analysing freeze-dried blood sera, Ali and coworkers investigated the use of ATR-FTIR spectroscopy for the differential diagnosis of dengue and hepatitis C infections in human blood. With the aid of multivariate analysis, the authors reported a classification accuracy of 99.2%, providing a viable alternative method for the differential diagnosis of viral diseases \u003csup\u003e10\u003c/sup\u003e. These recent findings demonstrate the suitability of ATR-FTIR spectroscopy coupled with multivariate analysis for analysing biological samples, in special blood plasma/sera, not only for identifying the presence or absence of diseases but also for the development of alternative methods for differential diagnosis. Therefore, the main purpose of this pilot study was to develop a simple, fast and low-cost methodology based on ATR-FTIR spectroscopy and multivariate classification for the accurate differential diagnosis of leptospirosis and dengue in humans via blood plasma analysis. Considering that both diseases present nonspecific acute symptoms that may lead to misinterpretation and incorrect diagnosis, there is still a need for alternative methods that can improve diagnosis and, consequently, the correct medical management of patients at initial phases, which contributes to decreasing both the mortality rates and the risks of outbreaks, especially in endemic regions.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\n \u003ch2\u003eCollection and preparation of specimens\u003c/h2\u003e\n \u003cp\u003eThis study involved patients living in Colombia who were admitted to different hospitals for medical assistance and treatment for symptoms of infection. All samples were collected and processed by the Laborat\u0026oacute;rio de Salud P\u0026uacute;blica of the Departamento del Atl\u0026aacute;ntico. All experiments were approved by the medical ethics committee at the Universidad del Atl\u0026aacute;ntico (Colombia), with an endorsement letter emitted in August 2023. Informed consent was obtained from each human participant and all methods were performed in accordance with the relevant guidelines and regulations (Declaration of Helsinki). Blood plasma samples from 114 patients in different phases of infection (n\u0026thinsp;=\u0026thinsp;43 for leptospirosis and n\u0026thinsp;=\u0026thinsp;71 for dengue), between men and women aged five to sixty years, were collected by venipuncture and transferred to tubes containing EDTA as an anticoagulant. Within 2 hours after collection, blood plasma samples were separated by density gradient, and aliquots were placed into cryogenic tubes and stored at -80\u0026deg;C until analysis. Leptospirosis was diagnosed by the reference serological test MAT (microscopic agglutination test) and ELISA for IgM. Dengue diagnosis was carried out by ELISA tests for IgM and NS1 (dengue antigen protein). According to the inclusion criteria, only samples collected within 18 months after blood collection were considered in this study.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\n \u003ch2\u003eATR-FTIR spectroscopy\u003c/h2\u003e\n \u003cp\u003eATR-FTIR spectral measurements were performed using a JASCO 4700 FTIR spectrometer (JASCO Corporation, Tokyo, Japan) with an ATR attachment containing a diamond crystal internal reflective element at a 45\u0026deg; incidence angle of the IR beam. The instrument was set up to perform a total of 16 scans with 4 cm\u003csup\u003e-1\u003c/sup\u003e spectral resolution on both the background and sample. The ATR crystal was cleaned with 98% isopropanol, and before each sample spectral acquisition, a background was collected. An aliquot of 20 \u0026micro;L of each room temperature blood plasma sample was placed onto aluminum foil as the substrate for spectral acquisition, according to Cui and colleagues \u003csup\u003e18\u003c/sup\u003e. Each sample was analysed in triplicate, resulting in a total of 342 spectra (n\u0026thinsp;=\u0026thinsp;129 for leptospirosis and n\u0026thinsp;=\u0026thinsp;213 for dengue).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\n \u003ch2\u003eData analysis and chemometric methods\u003c/h2\u003e\n \u003cp\u003eAll computational analyses, including importing, preprocessing and construction of multivariate classification models, were performed within the MATLAB\u0026reg; 2022 environment (MathWorks Inc., Natick, MA, USA) by using PLS Toolbox version 9.2 (Eigenvector Research Inc., USA) and laboratory-made routines. Initially, every plasma sample was analysed as its mean spectra, which were calculated using the three spectral replicates obtained for the sample, resulting in 43 mean spectra for leptospirosis samples and 71 mean spectra for dengue samples. Additionally, the ATR-FTIR mean spectra were preprocessed by selecting the fingerprint region between 1900 and 1000 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, resulting in a 114 x 935 matrix that was baseline corrected (automatic weighted least squares) and normalized to the amide I peak (i.e., ~ 1650 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e). For modelling, the samples were divided into training (~\u0026thinsp;70%, n\u0026thinsp;=\u0026thinsp;77), validation (~\u0026thinsp;15%, n\u0026thinsp;=\u0026thinsp;18) and prediction (~\u0026thinsp;15%, n\u0026thinsp;=\u0026thinsp;18) sets by applying the classical Kennard-Stone algorithm to the ATR-FTIR mean spectra \u003csup\u003e19\u003c/sup\u003e. One sample of the leptospirosis class was excluded and not modelled based on its outlier spectral behaviour. The training samples were used for model building and optimization (feature selection by the SPA and GA algorithms), the validation set was used to evaluate the performance of the models, and the prediction set was applied to test the optimized models in a real case situation when dealing with unknown samples.\u003c/p\u003e\n \u003cp\u003eThe discrimination models PCA-LDA, PCA-QDA, SPA-LDA, SPA-QDA, GA-LDA and GA-QDA were built by using laboratory-made routines for MATLAB software. A detailed description of the algorithms has already been provided by our research group and can be found elsewhere \u003csup\u003e20,21\u003c/sup\u003e. The GA-LDA/QDA calculations were performed for 40 generations with 80 chromosomes each. The one-point crossover and mutation probabilities were set to 60% and 10%, respectively. Moreover, the algorithm was repeated three times, starting from different random initial populations. The best solution (in terms of the fitness value) resulting from the three GA repetitions was employed.\u003c/p\u003e\n \u003cp\u003eCalculations of figures of merit, including sensitivity (probability that a test result will be positive when disease is present), specificity (probability that a test result will be negative when disease is not present), accuracy (number of samples correctly classified considering true and false negatives) and both the area under the ROC curve (AUC) and F Score, for indicating model performance, were obtained for this study as important quality measures to evaluate the discriminant ability of the models when applied to unknown/test samples. These parameters have a maximum value of 1 (100%) and a minimum of 0 and can be obtained by using the following equations \u003csup\u003e22,23\u003c/sup\u003e:\u003c/p\u003e\n \u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\n \u003cdiv class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e\u003cimg src=\"https://myfiles.space/user_files/122228_c8a1650c59388082/122228_custom_files/img1721030431.png\"\u003e\u003cbr\u003e\u003c/div\u003e\n \u003c/div\u003e\n \u003cdiv id=\"Equ5\" class=\"Equation\"\u003ewhere FN is defined as a false negative and FP is defined as a false positive. TP and TN are defined as true positive and true negative, respectively. Herein, for all calculations, leptospirosis samples were considered the positive class (\u0026ldquo;disease group\u0026rdquo;), and dengue samples were considered the negative class (\u0026ldquo;control group\u0026rdquo;).\u003c/div\u003e\n\u003c/div\u003e"},{"header":"Results","content":"\u003cp\u003eIn this study, a total of 114 blood plasma samples from patients suffering from acute symptoms of dengue (n\u0026thinsp;=\u0026thinsp;71) or leptospirosis (n\u0026thinsp;=\u0026thinsp;43) were analysed by infrared spectroscopy using the ATR mode, directly with no need for sample pretreatment. The raw spectra of all the samples are shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e-A in the fingerprint region between 1900 and 1000 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, where important biomolecules present significant absorption bands. Previously, for the construction of the discrimination models, spectra were preprocessed with baseline correction and normalization to the amide I peak to remove physical interferences and highlight chemical signals between the classes, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e-B. The mean preprocessed spectra of both classes, Dengue and Leptospirosis, are presented in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e-C, which shows strong similarity in their absorption bands.\u003c/p\u003e \u003cp\u003e[Insert Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA, B and C here]\u003c/p\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e shows the score plot for PCA of the dataset, with 95.67% of the explained variance accounting for the first two principal components. As the first attempt to observe possible clusters within samples, PCA analysis revealed no clear separation between samples belonging to the dengue or leptospirosis classes. On the other hand, there is an important observation in this graphic related to the presence of an outlier sample of the leptospirosis class that lies very far from the confidence limits. This sample was removed from the dataset before the construction of the discrimination models, resulting in a matrix of size 113 x 935.\u003c/p\u003e \u003cp\u003e[Insert Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e here]\u003c/p\u003e \u003cp\u003eFor modelling, samples were split into three sets for calibration, validation and prediction according to the calculations of the Kennard-Stone algorithm. Linear and quadratic discrimination models were constructed by using either the scores of principal components (PCA-LDA and PCA-QDA) or a subset of selected spectral variables obtained by SPA or GA algorithms (SPA-LDA, SPA-QDA, GA-LDA, GA-QDA). Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e represents the confusion matrix for each model, considering the samples of the prediction set.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003e\u003cb\u003e\u0026ndash;\u003c/b\u003e Confusion matrices for the discrimination models for dengue and leptospirosis using the prediction set.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"5\" nameend=\"c5\" namest=\"c1\"\u003e \u003cp\u003eModel Actual Class\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eDengue\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eLeptospirosis\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003ePCA-LDA\u003c/b\u003e (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e)\u003cb\u003e*\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDengue\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLeptospirosis\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eSPA-LDA\u003c/b\u003e (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDengue\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLeptospirosis\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eGA-LDA (33)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDengue\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLeptospirosis\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003ePCA-QDA\u003c/b\u003e (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDengue\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLeptospirosis\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eSPA-QDA\u003c/b\u003e (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDengue\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLeptospirosis\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eGA-QDA (31)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDengue\u003c/p\u003e \u003cp\u003eLeptospirosis\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e11\u003c/p\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0\u003c/p\u003e \u003cp\u003e7\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"5\"\u003e* Number of principal components or selected variables is shown in parentheses.\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e[Insert Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e here]\u003c/p\u003e \u003cp\u003eFor the LDA models, GA-LDA and SPA-LDA provided poor classification for the leptospirosis class, while PCA-LDA, which uses the scores of three principal components, misclassified only one sample in that class. On the other hand, for discriminating dengue patients, SPA and GA-LDA outperformed PCA-LDA. Quadratic discrimination did not enhance the classification performance of PCA-QDA for either of the two classes; however, it provided better results for SPA-QDA for leptospirosis samples. In the case of GA-QDA, an outstanding improvement in classification was observed for both classes, with the model reaching 100% correct classification for all prediction samples, using only 31 spectral variables.\u003c/p\u003e \u003cp\u003eTo provide a better evaluation of the model\u0026rsquo;s performance, some figures of merit, including sensitivity, specificity, accuracy, F-Score and AUC calculations, are depicted in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003e\u003cb\u003e\u0026ndash;\u003c/b\u003e Figures of merit for the discrimination models for dengue and leptospirosis using the prediction set.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eModel\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSensitivity (%)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSpecificity (%)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eAccuracy\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eF Score\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eAUC\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003ePCA-LDA\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e85.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e63.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e72.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.730\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.746\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eSPA-LDA\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e28.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e72.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e55.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.409\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.506\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eGA-LDA\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e57.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e81.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e72.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.672\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.694\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003ePCA-QDA\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e85.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e63.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e72.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.730\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.746\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eSPA-QDA\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e57.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e63.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e61.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.601\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.603\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eGA-QDA\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1.00\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e[Insert Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e here]\u003c/p\u003e \u003cp\u003eConsidering leptospirosis as the positive class, the sensitivity of the models (meaning their ability to correctly identify patients positive for leptospirosis infection) for the prediction set ranged from 28.5\u0026ndash;100%. The specificity (i.e., the ability of the models to identify patients positive for dengue infection) varied from 63.6\u0026ndash;100%. The accuracy, which reflects the performance of the discrimination models, ranged from 55.5\u0026ndash;100%. The best model for linear discrimination was PCA-LDA, with an accuracy of 72.2% and F- score and AUC of 0.730 and 0.746, respectively. When using QDA, the GA-QDA model presented remarkable quality parameters for classification, reaching maximum rates for all the metrics calculated herein.\u003c/p\u003e \u003cp\u003eThe discriminant function showing the separation between leptospirosis and dengue samples is depicted in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e-A for the best model, GA-QDA. The selected subset of 31 out of 935 variables is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e-B, highlighting the main spectral regions used for correct discrimination. The area under the ROC curve is plotted in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e-C for both the GA-LDA and GA-QDA models to demonstrate the improvement in the performance of the GA model when the quadratic discrimination function was applied.\u003c/p\u003e \u003cp\u003e[Insert Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e-A, B and C here]\u003c/p\u003e \u003cp\u003eAnalysis of variance was applied to the 31 variables chosen by the GA-QDA model to verify significant differences between the means of the absorption signals of the dengue and leptospirosis samples. At a confidence level of 95%, a set of 10 out of 31 variables presented significant differences (P\u0026thinsp;\u0026lt;\u0026thinsp;0.05) in the ANOVA calculations, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e4\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e[Insert Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e4\u003c/span\u003e here]\u003c/p\u003e \u003cp\u003eAccording to the box plots, the mean absorbance values of the dengue samples were greater than those of the leptospirosis samples for all the variables except for the wavenumber at 1612 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. Additionally, more outliers were identified for the Dengue class, while for Leptospirosis, a more evident spread was observed, with larger boxes as well as whiskers in a larger range, excluding the wavenumber at 1296 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eLeptospirosis and dengue are both infectious diseases that present very similar acute initial symptoms and usually include high-grade fever, strong headache, nausea, chills and myalgia. Such similarity in the acute phase may result in additional risks for patients, considering that the differential diagnosis of both illnesses may be challenging and that medical care should be specific for each disease since they are viral (dengue) and bacterial (leptospirosis) infections. A misleading diagnosis not only worsens the prognosis since these infections can evolve rapidly for severe conditions but also impairs control and mitigation in endemic areas.\u003c/p\u003e \u003cp\u003eThe use of spectroscopic techniques to analyse biological samples, known as biospectroscopy, has gained great attention in recent decades since it allows the investigation of diseases based on biochemical changes that occur as a consequence of pathological events \u003csup\u003e24\u003c/sup\u003e. Vibrational spectroscopies have been successfully applied in many studies in medical chemistry \u003csup\u003e25\u003c/sup\u003e since they provide alternative methods for screening, identification and diagnosis of diseases through simple, cost-effective and accurate methodologies. Herein, we used different chemometric methods to analyse ATR-FTIR data for the differential diagnosis of leptospirosis and dengue in patients from Colombia, an endemic country for both diseases. With the exploratory analysis of PCA, no clear separation was observed in the samples; however, an outlier could be visually identified. This unique sample refers to a nonpathogenic strain of \u003cem\u003eLeptospira\u003c/em\u003e known as saprophytic, as confirmed by the reference MAT test. This finding demonstrates the usefulness of PCA for identifying patterns and improving comprehension of the dataset.\u003c/p\u003e \u003cp\u003eClassification models based on linear or quadratic discriminant analysis were constructed to allow the differential diagnosis of leptospirosis from dengue. With LDA, the best accuracy obtained was 72.2% by the PCA-LDA and GA-LDA models, representing an unsatisfactory ability to differentiate the diseases. Quadratic discrimination by the GA-QDA model considerably improved the classification results, achieving 100% accuracy, sensitivity and specificity, which allows the differential diagnosis of leptospirosis and dengue. These results highlight that when analysing imbalanced datasets (both classes do not have the same number of samples), quadratic discrimination should be the best suited tool. The 31 main variables responsible for discriminating the two classes, selected by GA, were compared by using analysis of variance. The ANOVA results, at a confidence level of 95%, showed that only 10 variables (1296 cm\u003csup\u003e-1\u003c/sup\u003e, 1612 cm\u003csup\u003e-1\u003c/sup\u003e, 1673 cm\u003csup\u003e-1\u003c/sup\u003e, 1677 cm\u003csup\u003e-1\u003c/sup\u003e, 1678 cm\u003csup\u003e-1\u003c/sup\u003e, 1689 cm\u003csup\u003e-1\u003c/sup\u003e, 1694 cm\u003csup\u003e-1\u003c/sup\u003e, 1711 cm\u003csup\u003e-1\u003c/sup\u003e, 1713 cm\u003csup\u003e-1\u003c/sup\u003e and 1719 cm\u003csup\u003e-1\u003c/sup\u003e) presented significant differences in absorbance between the leptospirosis and dengue classes. These specific wavenumbers represent the most useful spectral information accounting for the biochemical changes that mark a specific infection. Some molecular bonds can be tentatively assigned to the ten significant variables, such as the band of amide III (~\u0026thinsp;from 1290 to 1334 cm\u003csup\u003e-1\u003c/sup\u003e), the bands of amide I (C\u0026thinsp;=\u0026thinsp;N associated with β-sheets, random coil and C\u0026thinsp;=\u0026thinsp;O stretch of conjugated ketones) in the region of approximately 1620 to 1690 cm\u003csup\u003e-1\u003c/sup\u003e, and the region from 1700 to 1720 related to the C\u0026thinsp;=\u0026thinsp;O stretching vibrations of ketones and aldehydes \u003csup\u003e26,27\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe differential diagnosis between leptospirosis and dengue by infrared spectroscopy and multivariate analysis is based on the biochemical variations that occur as a consequence of the pathological process. The spectral regions responsible for the discrimination of samples cannot be directly associated with specific molecules, such as biomarkers; however, they represent important information about the classes of biomolecules most affected by these illnesses. Therefore, other techniques, such as mass spectrometry, should be more suitable for the investigation of possible biomarkers, and a larger number of samples should be included in further studies. Nonetheless, the spectroscopic-multivariate method presented herein provides a very useful alternative for overcoming the challenges related to leptospirosis and dengue infections as a simple, cost-effective and accurate methodology for differential diagnosis.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eHerein, we demonstrated the use of ATR-FTIR spectroscopy combined with multivariate analysis for the differential diagnosis of two infectious diseases, leptospirosis and dengue. By quadratic discriminant analysis, the GA-QDA model reached maximum rates of correct classification, with sensitivity, specificity and accuracy values of 100% when analysing test (unknown) samples. The outstanding results obtained in this pilot study highlight the capability of this methodology to be applied in clinical practice to serve as a rapid and accurate test for discriminating between the two illnesses.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/strong\u003eThe authors of this work would like to thank the Chemistry Postgraduation Program of the Federal University of Rio Grande do Norte.\u0026nbsp;K\u0026aacute;ssio M.G. Lima thanks the National Council for Scientific and Technological Development grant (Ref. 305562/2020-7).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets used and/or analysed during the current study available from the corresponding author on reasonable request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eRodr\u0026iacute;guez-Salazar, C. A. \u003cem\u003eet al.\u003c/em\u003e Manifestaciones cl\u0026iacute;nicas y hallazgos de laboratorio de una serie de casos febriles agudos con diagn\u0026oacute;stico presuntivo de infecci\u0026oacute;n por el virus dengue. Quind\u0026iacute;o (Colombia). \u003cem\u003eInfectio\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, 84\u0026ndash;92 (2016).\u003c/li\u003e\n\u003cli\u003eHartskeerl, R. A., Collares-Pereira, M. \u0026amp; Ellis, W. A. Emergence, control and re-emerging leptospirosis: Dynamics of infection in the changing world. \u003cem\u003eClinical Microbiology and Infection\u003c/em\u003e vol. 17 494\u0026ndash;501 Preprint at https://doi.org/10.1111/j.1469-0691.2011.03474.x (2011).\u003c/li\u003e\n\u003cli\u003eClemen G, Angel J, Montes C, Jr, T. \u0026amp; Osorio L. \u003cem\u003eContribuci\u0026oacute;n de La Prueba R\u0026aacute;pida NS1 e IgM al Diagn\u0026oacute;stico de Dengue En Colombia En El Periodo Pre-Zika\u003c/em\u003e.\u003c/li\u003e\n\u003cli\u003ePeeling, R. W. \u003cem\u003eet al.\u003c/em\u003e Evaluation of diagnostic tests: Dengue. \u003cem\u003eNat Rev Microbiol\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, S30\u0026ndash;S38 (2010).\u003c/li\u003e\n\u003cli\u003eG\u0026aacute;rcia, J. D., Ag\u0026ucirc;ero, J., Parra, JA. \u0026amp; Santos, MF. Enfermedades infecciosas. Concepto. Clasificaci\u0026oacute;n. Aspectos generales y espec\u0026iacute;ficos de las infecciones. Criterios de sospecha de enfermedad infecciosa. Pruebas diagn\u0026oacute;sticas complementarias. Criterios de indicaci\u0026oacute;n. \u003cem\u003eMedicina - Programa de formaci\u0026ograve;n Medica Continuada acreditado\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 3251\u0026ndash;3264 (2010).\u003c/li\u003e\n\u003cli\u003ePinto, G. V. \u003cem\u003eet al.\u003c/em\u003e Current methods for the diagnosis of leptospirosis: Issues and challenges. \u003cem\u003eJournal of Microbiological Methods\u003c/em\u003e vol. 195 Preprint at https://doi.org/10.1016/j.mimet.2022.106438 (2022).\u003c/li\u003e\n\u003cli\u003eSuwannin, P. \u003cem\u003eet al.\u003c/em\u003e Enhancing leptospirosis control with nanosensing technology: A critical analysis. \u003cem\u003eComparative Immunology, Microbiology and Infectious Diseases\u003c/em\u003e vol. 104 Preprint at https://doi.org/10.1016/j.cimid.2023.102092 (2024).\u003c/li\u003e\n\u003cli\u003eSantos, M. C. D. \u003cem\u003eet al.\u003c/em\u003e ATR-FTIR spectroscopy with chemometric algorithms of multivariate classification in the discrimination between healthy: Vs. dengue vs. chikungunya vs. zika clinical samples. \u003cem\u003eAnalytical Methods\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 1280\u0026ndash;1285 (2018).\u003c/li\u003e\n\u003cli\u003eNaseer, K., Ali, S. \u0026amp; Qazi, J. ATR-FTIR spectroscopy based differentiation of typhoid and dengue fever in infected human sera. \u003cem\u003eInfrared Phys Technol\u003c/em\u003e \u003cstrong\u003e114\u003c/strong\u003e, (2021).\u003c/li\u003e\n\u003cli\u003eAli, Salmann., Naseer, Khulla., Hussain, Irshad. \u0026amp; Qazi, Javaria. ATR-FTIR spectroscopy-based differentiation of hepatitis C and dengue infection in human freeze-dried sera. \u003cem\u003eInfrared Phys Technol\u003c/em\u003e \u003cstrong\u003e118\u003c/strong\u003e, 1350\u0026ndash;4495 (2021).\u003c/li\u003e\n\u003cli\u003eMorais, C. L. M., Lima, K. M. G., Singh, M. \u0026amp; Martin, F. L. Tutorial: multivariate classification for vibrational spectroscopy in biological samples. \u003cem\u003eNature Protocols\u003c/em\u003e vol. 15 2143\u0026ndash;2162 Preprint at https://doi.org/10.1038/s41596-020-0322-8 (2020).\u003c/li\u003e\n\u003cli\u003eSantos, M. C. D. \u003cem\u003eet al.\u003c/em\u003e ATR-FTIR spectroscopy with chemometric algorithms of multivariate classification in the discrimination between healthy: Vs. dengue vs. chikungunya vs. zika clinical samples. \u003cem\u003eAnalytical Methods\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 1280\u0026ndash;1285 (2018).\u003c/li\u003e\n\u003cli\u003eKawakami, R. \u003cem\u003eet al.\u003c/em\u003e \u003cem\u003eAspects of the Successive Projections Algorithm for Variable Selection in Multivariate Calibration Applied to Plasma Emission Spectrometry\u003c/em\u003e. \u003cem\u003eAnalytica Chimica Acta\u003c/em\u003e vol. 443 (2001).\u003c/li\u003e\n\u003cli\u003eSilva, H. K. T. de A. \u003cem\u003eet al.\u003c/em\u003e Detection of terbufos in cases of intoxication by means of entomotoxicological analysis using ATR-FTIR spectroscopy combined with chemometrics. \u003cem\u003eActa Trop\u003c/em\u003e \u003cstrong\u003e238\u003c/strong\u003e, (2023).\u003c/li\u003e\n\u003cli\u003eDixon, S. J. \u0026amp; Brereton, R. G. Comparison of performance of five common classifiers represented as boundary methods: Euclidean Distance to Centroids, Linear Discriminant Analysis, Quadratic Discriminant Analysis, Learning Vector Quantization and Support Vector Machines, as dependent on data structure. \u003cem\u003eChemometrics and Intelligent Laboratory Systems\u003c/em\u003e \u003cstrong\u003e95\u003c/strong\u003e, 1\u0026ndash;17 (2009).\u003c/li\u003e\n\u003cli\u003eMorais, C. L. M. \u0026amp; Lima, K. M. G. Principal component analysis with linear and quadratic discriminant analysis for identification of cancer samples based on mass spectrometry. \u003cem\u003eJ Braz Chem Soc\u003c/em\u003e \u003cstrong\u003e29\u003c/strong\u003e, 472\u0026ndash;481 (2018).\u003c/li\u003e\n\u003cli\u003eSalmann, A., Khulla, N., Irshad, H. \u0026amp; Javaria, Q. ATR-FTIR spectroscopy-based differentiation of hepatitis C and dengue infection in human freeze-dried sera. \u003cem\u003eInfrared Physics and Technology \u003c/em\u003e1\u0026ndash;4 (2021).\u003c/li\u003e\n\u003cli\u003eCui, L., Butler, H. J., Martin-Hirsch, P. L. \u0026amp; Martin, F. L. Aluminium foil as a potential substrate for ATR-FTIR, transflection FTIR or Raman spectrochemical analysis of biological specimens. \u003cem\u003eAnalytical Methods\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 481\u0026ndash;487 (2016).\u003c/li\u003e\n\u003cli\u003eKennard, R. W. \u0026amp; Stone, L. A. \u003cem\u003eComputer Aided Design of Experiments\u003c/em\u003e. vol. 11 http://www.jstor.org/stable/1266770 . (1969).\u003c/li\u003e\n\u003cli\u003eMorais, C. L. M., Lima, K. M. G., Singh, M. \u0026amp; Martin, F. L. Tutorial: multivariate classification for vibrational spectroscopy in biological samples. \u003cem\u003eNature Protocols\u003c/em\u003e vol. 15 2143\u0026ndash;2162 Preprint at https://doi.org/10.1038/s41596-020-0322-8 (2020).\u003c/li\u003e\n\u003cli\u003eNeves, A. C. O. \u003cem\u003eet al.\u003c/em\u003e ATR-FTIR and multivariate analysis as a screening tool for cervical cancer in women from northeast Brazil: A biospectroscopic approach. \u003cem\u003eRSC Adv\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 99648\u0026ndash;99655 (2016).\u003c/li\u003e\n\u003cli\u003eMorais, C. L. M., Lima, K. M. G., Singh, M. \u0026amp; Martin, F. L. Tutorial: multivariate classification for vibrational spectroscopy in biological samples. \u003cem\u003eNature Protocols\u003c/em\u003e vol. 15 2143\u0026ndash;2162 Preprint at https://doi.org/10.1038/s41596-020-0322-8 (2020).\u003c/li\u003e\n\u003cli\u003eL\u0026oacute;pez, V., Fern\u0026aacute;ndez, A., Garc\u0026iacute;a, S., Palade, V. \u0026amp; Herrera, F. An insight into classification with imbalanced data: Empirical results and current trends on using data intrinsic characteristics. \u003cem\u003eInf Sci (N Y)\u003c/em\u003e \u003cstrong\u003e250\u003c/strong\u003e, 113\u0026ndash;141 (2013).\u003c/li\u003e\n\u003cli\u003eda Silva, T. G. \u003cem\u003eet al.\u003c/em\u003e Spectrochemical analysis of blood combined with chemometric techniques for detecting osteosarcopenia. \u003cem\u003eSci Rep\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, (2023).\u003c/li\u003e\n\u003cli\u003eAlkhuder, K. Fourier transform infrared spectroscopy: A universal optical sensing technique with auspicious application prospects in the diagnosis and management of autoimmune diseases. \u003cem\u003ePhotodiagnosis and Photodynamic Therapy\u003c/em\u003e vol. 42 Preprint at https://doi.org/10.1016/j.pdpdt.2023.103606 (2023).\u003c/li\u003e\n\u003cli\u003eWalsh, M. J. \u003cem\u003eet al.\u003c/em\u003e IR microspectroscopy: potential applications in cervical cancer screening. \u003cem\u003eCancer Letters\u003c/em\u003e vol. 246 1\u0026ndash;11 Preprint at https://doi.org/10.1016/j.canlet.2006.03.019 (2007).\u003c/li\u003e\n\u003cli\u003ePurandare, N. C. \u003cem\u003eet al.\u003c/em\u003e Infrared spectroscopy with multivariate analysis segregates low-grade cervical cytology based on likelihood to regress, remain static or progress. \u003cem\u003eAnalytical Methods\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 4576\u0026ndash;4584 (2014).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"leptospirosis, dengue, differential diagnosis, ATR-FTIR, multivariate analysis","lastPublishedDoi":"10.21203/rs.3.rs-4607844/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4607844/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eDengue and leptospirosis are considered systemic and very dynamic illnesses in which a patient can rapidly progress from mild to severe conditions. Both diseases present very similar acute initial symptoms, a fact that may result in a challenging differential diagnosis at the initial phases. Herein, we present the application of attenuated total reflectance-Fourier transform infrared (ATR-FTIR) spectroscopy combined with multivariate analysis to perform differential diagnosis of leptospirosis and dengue by analysing blood plasma. The spectra of 114 samples from patients in different phases of infection (\u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;43 for leptospirosis and \u003cem\u003en\u003c/em\u003e\u0026thinsp;=\u0026thinsp;71 for dengue) were analysed by either linear or quadratic discrimination in association with genetic algorithms, successive projection algorithms and principal component analysis for feature selection/extraction. The best model, GA-QDA, achieved outstanding results in terms of maximum (100%) sensitivity, specificity and accuracy for classifying both classes by using only 31 spectral variables. The ANOVA calculations, at a confidence level of 95%, highlighted a set of 10 variables selected by the GA-QDA model (1296 cm\u003csup\u003e-1\u003c/sup\u003e, 1612 cm\u003csup\u003e-1\u003c/sup\u003e, 1673 cm\u003csup\u003e-1\u003c/sup\u003e, 1677 cm\u003csup\u003e-1\u003c/sup\u003e, 1678 cm\u003csup\u003e-1\u003c/sup\u003e, 1689 cm\u003csup\u003e-1\u003c/sup\u003e, 1694 cm\u003csup\u003e-1\u003c/sup\u003e, 1711 cm\u003csup\u003e-1\u003c/sup\u003e, 1713 cm\u003csup\u003e-1\u003c/sup\u003e and 1719 cm\u003csup\u003e-1\u003c/sup\u003e) with significant differences in the absorbance means between the Leptospirosis and Dengue classes. These specific wavenumbers represent the most useful spectral information accounting for the biochemical changes that mark a specific infection. These remarkable results obtained in this pilot study highlight the viability of this methodology to be applied in clinical practice to serve as a simple and accurate test for discriminating between the two illnesses.\u003c/p\u003e","manuscriptTitle":"Infrared spectroscopy and multivariate analysis applied to the differential diagnosis of leptospirosis and dengue","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-07-18 20:47:47","doi":"10.21203/rs.3.rs-4607844/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"115a357b-4d02-4b30-b8ef-dda57fc8f0dd","owner":[],"postedDate":"July 18th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2024-08-01T11:24:39+00:00","versionOfRecord":[],"versionCreatedAt":"2024-07-18 20:47:47","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-4607844","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4607844","identity":"rs-4607844","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}