A computer vision chemometric-assisted approach to access pH and glucose influence on susceptibility of Candida pathogenic strains.

Microorganisms adapt to environmental conditions as a survival strategy for different interactions with the environment. The adaptive capacity of fungi allows them to cause disease at various sites of infection in humans. In this study, we propose digital images as responses of a complete factorial 23. Furthermore, we compared two experimental approaches: the experimental design (3D) and the checkerboard assay (2D) to know the influence of pH, glucose, and fluconazole concentration on different strains of the genus Candida. The digital images obtained from the factorial 23 were used as input in the PCA-ANOVA to analyze the results of this experimental design. pH modification in the culture medium modifies the susceptibility in some species less adapted to this type of modification. For the first time, to the best of our knowledge, digital images were used as input to PCA-ANOVA to obtain information on Candida spp.. Therefore, a higher concentration of antifungals is needed to inhibit the same strain at a lower pH. In short, we present an alternative with less use of reagents and time. In addition, the use of digital images allows obtaining information about fungal susceptibility with three or more factors.