SlideImaging

The SlideImaging project demonstrated the feasibility of automation in the acquisition, management, and diagnosis of Indirect Immunofluorescence (IFI) images for the diagnosis of systemic autoimmune diseases. Despite the emergence of new systems for autoantibody detection, the IFI method remains the gold standard for antinuclear antibody (ANA) detection. The IFI method is time-consuming, with high inter- and intra-operator variability, and requires expert personnel to read and interpret the slides. Our project, especially in its initial phase, significantly contributed to raising awareness among the scientific community regarding the feasibility of automation, including in the IFI reading phase, leading to the implementation of an automatic fluorescence intensity and pattern classifier (CAD).

Over the years, we have expanded the panel of recognizable antigen-antibody specificities, expanding the typology of immunofluorescence patterns that can be analyzed, and we have built and tested algorithms that allow us to automatically identify which regions of the preparation the classifier should analyze to implement automatic diagnosis.

In parallel, we built a classifier for the automation of anti-dsDNA in IIF, using Critidia Luciliae as a substrate. The data obtained showed excellent performance of the automated system compared to reading by an "expert" operator.

In the subsequent phases of the project, we applied the automatic diagnosis obtained through CAD to a larger number of cases, involving three additional Italian centers of reference in the field. The data obtained further validated the preliminary findings.

Project start date

2006