Fig 1.
Morphological classifications within a sample titration.
(A) HAI dilution series displaying a typical transition point in the absence of non-specific inhibition; a direct comparison of the first two wells to the negative control in the far right well shows very similar size, shape and overall morphology of the buttons. (B) HAI dilution series displaying non-uniform morphologies due to non-specific inhibition within the first well—an endpoint is reached in the third well; a direct comparison of the third well to the negative control in the far right well shows similar size, shape and overall morphology. (C) HAI dilution series displaying non-uniform morphologies due to non-specific inhibition; a direct comparison of the negative control in the far right well highlights the different morphology of the buttons as a function of dilution. (D) HAI dilution series displaying wells which are all fully agglutinated and have no inhibition as compared to the negative control in the far right well.
Table 1.
Sample classifications based on morphological appearances.
Fig 2.
Screen shot of Cypher One user interface.
Example HAI plate image within a screen shot of the Cypher One software application. High resolution digital images were captured and titer values determined using an image analysis algorithm within the software. The experimental information, image and resulting titer values are stored as digital records within the system database.
Fig 3.
Histogram representing the number of dilutions different between the Cypher One and human titer interpretation for the entire dataset (n = 2200).
Table 2.
Cypher One agreement to expert human reader.
Table 3.
Cypher One agreement to expert human reader per sample classification.
Fig 4.
Log titer correlation plot for the guinea pig and turkey RBC datasets.
(A) Shows the log titer correlation plot between interpretation methods for each assigned titer value for the guinea pig dataset. (B) Shows the log titer correlation plot between interpretation methods for each assigned titer value for the turkey dataset. The range of log titer values shown on the X axis reflect the titer values assigned by the human reader. The average Cypher One interpretation at all titer values is represented on the Y axis. Assigned titer values ranged from 5 to 2560, or log titer values of 0.7 to 3.4, respectively. Log titer values of 0.7 (HAI titer of 5) exhibited either non-specific inhibition or full agglutination across the entire dilution series. The number shown above each data point indicates the number of samples included in each data point.