Figure 1.
A: CSA consists of 80 different chemically responsive nanoporous pigments. B: Examples of four classes of chemically-responsive dyes: (i) metal-ion-containing dyes that respond to Lewis basicity, (ii) pH indicators that respond to Brønsted acidity/basicity, (iii) dyes with large permanent dipoles (e.g., solvatochromic dyes) that respond to local polarity, and (iv) redox indicators that respond to electrochemical reaction.
Figure 2.
Experimental setup for solid media study.
A: An incubator hosting the scanner. B: The scanner holding six Petri dishes housed within a closed plastic box. All experiments were performed using standard safety protocols in an ACDP Containment Level 3 laboratory.
Table 1.
Bacteria detection time based on strain and inoculum concentrations.
Figure 3.
Color difference maps of B. anthracis Ames, Y. pestis CO92, Y. pseudotuberculosis, and E. coli.
Each color represents the difference between the indicator color intensity measured before exposure and the intensity measured at the indicated detection time for each species. For visualization purposes, color difference maps are expanded from 4 to 8 bits per color (RGB range of 0–15 expanded to 0–255) [33].
Figure 4.
Species time response profiles.
Selected time response profiles of four different bacterial species at both low and high concentrations. At least six trials were collected per species.
Figure 5.
Species concentration trajectories in LDA space.
Trajectories for each species from low to high concentration are separable in LDA space, suggesting that species can be identified independent of inoculum concentration. Six control trials (black) are located at the origin.
Table 2.
Species discrimination between 4 different pathogenic bacteria plus a control (95% confidence intervals in parentheses).
Figure 6.
Strain time response profiles at high concentrations.
For each trial, red, green, and blue lines plot either the percent change in each color from its initial intensity or the rate of the color change. Trial duration varied depending on species growth rate.
Figure 7.
Strain separation in PCA space.
A: PCA plot separating three different strains of B. anthracis. B: Expansion showing the separation among different strains at low inoculum concentrations. C: PCA plot for two different strains of Y. pestis.
Table 3.
Strain discrimination within B. anthracis and Y. pestis. (95% confidence intervals in parentheses).
Table 4.
Confusion matrices of observed labels (row) and predicted labels (column) for species discrimination.
Table 5.
Confusion matrices of observed labels (row) and predicted labels (column) for B. anthracis strain discrimination.
Table 6.
Confusion matrices of observed labels (row) and predicted labels (column) for Y. pestis strain discrimination.