The spread of SARS-CoV-2, which causes the disease COVID-19, is difficult to control as some positive individuals, capable of transmitting the disease, can be asymptomatic. Thus, it remains critical to generate noninvasive, inexpensive COVID-19 screening systems. Two such methods include detection canines and analytical instrumentation, both of which detect volatile organic compounds associated with SARS-CoV-2. In this study, the performance of trained detection dogs is compared to a noninvasive headspace-solid phase microextraction-gas chromatography-mass spectrometry (HS-SPME-GC-MS) approach to identifying COVID-19 positive individuals. Five dogs were trained to detect the odor signature associated with COVID-19. They varied in performance, with the two highest-performing dogs averaging 88% sensitivity and 95% specificity over five double-blind tests. The three lowest-performing dogs averaged 46% sensitivity and 87% specificity. The optimized linear discriminant analysis (LDA) model, developed using HS-SPME-GC-MS, displayed a 100% true positive rate and a 100% true negative rate using leave-one-out cross-validation. However, the non-optimized LDA model displayed difficulty in categorizing animal hair-contaminated samples, while animal hair did not impact the dogs’ performance. In conclusion, the HS-SPME-GC-MS approach for noninvasive COVID-19 detection more accurately discriminated between COVID-19 positive and COVID-19 negative samples; however, dogs performed better than the computational model when non-ideal samples were presented.

The adaptable nature of the SARS-CoV-2 virus has led to the emergence of multiple viral variants of concern. This research builds upon a previous demonstration of sampling human hand odor to distinguish SARS-CoV-2 infection status in order to incorporate considerations of the disease variants. This study demonstrates the ability of human odor expression to be implemented as a non-invasive medium for the differentiation of SARS-CoV-2 variants. Volatile organic compounds (VOCs) were extracted from SARS-CoV-2-positive samples using solid phase microextraction (SPME) coupled with gas chromatography–mass spectrometry (GC–MS). Sparse partial least squares discriminant analysis (sPLS-DA) modeling revealed that supervised machine learning could be used to predict the variant identity of a sample using VOC expression alone. The class discrimination of Delta and Omicron BA.5 variant samples was performed with 95.2% (±0.4) accuracy. Omicron BA.2 and Omicron BA.5 variants were correctly classified with 78.5% (±0.8) accuracy. Lastly, Delta and Omicron BA.2 samples were assigned with 71.2% (±1.0) accuracy. This work builds upon the framework of non-invasive techniques producing diagnostics through the analysis of human odor expression, all in support of public health monitoring.

Since the beginning of the COVID-19 pandemic, there has been enormous interest in the development of measures that would allow for the swift detection of the disease. The rapid screening and preliminary diagnosis of SARS-CoV-2 infection allow for the instant identification of possibly infected individuals and the subsequent mitigation of the disease spread. Herein, the detection of SARS-CoV-2-infected individuals was explored using noninvasive sampling and low-preparatory-work analytical instrumentation. Hand odor samples were obtained from SARS-CoV-2-positive and -negative individuals. The volatile organic compounds (VOCs) were extracted from the collected hand odor samples using solid phase microextraction (SPME) and analyzed using gas chromatography coupled with mass spectrometry (GC-MS). Sparse partial least squares discriminant analysis (sPLS-DA) was used to develop predictive models using the suspected variant sample subsets. The developed sPLS-DA models performed moderately (75.8% (±0.4) accuracy, 81.8% sensitivity, 69.7% specificity) at distinguishing between SARS-CoV-2-positive and negative -individuals based on the VOC signatures alone. Potential markers for distinguishing between infection statuses were preliminarily acquired using this multivariate data analysis. This work highlights the potential of using odor signatures as a diagnostic tool and sets the groundwork for the optimization of other rapid screening sensors such as e-noses or detection canines.