The blood diagnosis of diabetes mellitus (DM) is highly accurate; however, it is an invasive, high-cost, and painful procedure. In this context, the combination of ATR-FTIR spectroscopy and machine learning techniques in other biological samples has been used as an alternative tool to develop a non-invasive, fast, inexpensive, and label-free diagnostic or screening platform for several diseases, including DM. In this study, we used the ATR-FTIR tool associated with linear discriminant analysis (LDA) and a support vector machine (SVM) classifier in order to identify changes in salivary components to be used as alternative biomarkers for the diagnosis of type 2 DM. The band area values of 2962 cm⁻¹, 1641 cm⁻¹, and 1073 cm⁻¹ were higher in type 2 diabetic patients than in non-diabetic subjects. The best classification of salivary infrared spectra was by SVM, showing a sensitivity of 93.3% (42/45), specificity of 74% (17/23), and accuracy of 87% between non-diabetic subjects and uncontrolled type 2 DM patients. The SHAP features of infrared spectra indicate the main salivary vibrational modes of lipids and proteins that are responsible for discriminating DM patients. In summary, these data highlight the potential of ATR-FTIR platforms coupled with machine learning as a reagent-free, non-invasive, and highly sensitive tool for screening and monitoring diabetic patients.

Developing affordable, rapid, and accurate biosensors is essential for SARS-CoV-2 surveillance and early detection. We created a bio-inspired peptide, using the SAGAPEP AI platform, for COVID-19 salivary diagnostics via a portable electrochemical device coupled to Machine Learning algorithms. SAGAPEP enabled molecular docking simulations against the SARS-CoV-2 Spike protein’s RBD, leading to the synthesis of Bio-Inspired Artificial Intelligence Peptide 1 (BIAI1). Molecular docking was used to confirm interactions between BIAI1 and SARS-CoV-2, and BIAI1 was functionalized on rhodamine-modified electrodes. Cyclic voltammetry (CV) using a [Fe(CN)₆]^3−/4 solution detected virus levels in saliva samples with and without SARS-CoV-2. Support vector machine (SVM)-based machine learning analyzed electrochemical data, enhancing sensitivity and specificity. Molecular docking revealed stable hydrogen bonds and electrostatic interactions with RBD, showing an average affinity of −250 kcal/mol. Our biosensor achieved 100% sensitivity, 80% specificity, and 90% accuracy for 1.8 × 10⁴ focus-forming units in infected saliva. Validation with COVID-19-positive and -negative samples using a neural network showed 90% sensitivity, specificity, and accuracy. This BIAI1-based electrochemical biosensor, integrated with machine learning, demonstrates a promising non-invasive, portable solution for COVID-19 screening and detection in saliva.

Data classification is a common task, which can be performed by both computers and human beings. However, a fundamental difference between them can be observed: computer-based classification considers only physical features (e.g., similarity, distance, or distribution) of input data; by contrast, brain-based classification takes into account not only physical features, but also the organizational structure of data. In this paper, we figure out the data organizational structure for classification using complex networks constructed from training data. Specifically, an unlabeled instance is classified by the importance concept characterized by Google's PageRank measure of the underlying data networks. Before a test data instance is classified, a network is constructed from vector-based data set and the test instance is inserted into the network in a proper manner. To this end, we also propose a measure, called spatio-structural differential efficiency, to combine the physical and topological features of the input data. Such a method allows for the classification technique to capture a variety of data patterns using the unique importance measure. Extensive experiments demonstrate that the proposed technique has promising predictive performance on the detection of heart abnormalities.

Zika virus (ZIKV) diagnosis is currently performed through an invasive, painful, and costly procedure using molecular biology. Consequently, the search for a non-invasive, more cost-effective, reagent-free, and sustainable method for ZIKV diagnosis is of great relevance. It is critical to prepare a global strategy for the next ZIKV outbreak given its devastating consequences, particularly in pregnant women. Attenuated total reflection–Fourier transform infrared (ATR-FTIR) spectroscopy has been used to discriminate systemic diseases using saliva; however, the salivary diagnostic application in viral diseases is unknown. To test this hypothesis, we intradermally challenged interferon-gamma gene knockout C57/BL6 mice with ZIKV (50 µL,105 FFU, n = 7) or vehicle (50 µL, n = 8). Saliva samples were collected on day three (due to the peak of viremia) and the spleen was also harvested. Changes in the salivary spectral profile were analyzed by Student’s t test (p < 0.05), multivariate analysis, and the diagnostic capacity by ROC curve. ZIKV infection was confirmed by real-time PCR of the spleen sample. The infrared spectroscopy coupled with univariate analysis suggested the vibrational mode at 1547 cm⁻¹ as a potential candidate to discriminate ZIKV and control salivary samples. Three PCs explained 93.2% of the cumulative variance in PCA analysis and the spectrochemical analysis with LDA achieved an accuracy of 93.3%, with a specificity of 87.5% and sensitivity of 100%. The LDA-SVM analysis showed 100% discrimination between both classes. Our results suggest that ATR-FTIR applied to saliva might have high accuracy in ZIKV diagnosis with potential as a non-invasive and cost-effective diagnostic tool.