Comparison of Anti-dsDNA Evaluation with 20x and 40x Objectives in Automated Systems to Expert Agreement

BACKGROUND: In this study, we investigated the effect of using a 40x objective instead of the recommended 20x objective in the computer-assisted automated evaluation of anti-dsDNA antibodies using the Crithidia luciliae indirect immunofluorescence (CLIF) test. By using a 40x objective, we aimed to increase the accuracy and specificity of the CLIF test by improving image clarity and enabling easier interpretation by the expert physician. METHODS: Anti-dsDNA tests of 156 positive and 40 negative samples were evaluated using the automated EURO-Pattern system at 20x and 40x magnifications. The results were compared with the assessments of experienced physicians. Statistical analysis included chi-squared and kappa agreement tests. Sensitivity, specificity, and accuracy metrics were calculated for both objectives. RESULTS: When evaluated with the 20x objective, the system achieved a sensitivity of 100%, specificity of 29.85%, and accuracy of 50.26%, with a kappa coefficient of 0.199 (95% CI: 0.129 - 0.267). With the 40x objective, sensitivity was 94.55%, specificity was 90.30%, and accuracy was 91.53%, with a kappa coefficient of 0.805 (95% CI: 0.714 - 0.895). The agreement between the automated system and the expert evaluations significantly improved with the 40x objective. CONCLUSIONS: Using a 40x microscope objective enhances the compatibility between automated systems and expert evaluations, providing clearer and larger images. This adjustment reduces false positives, increases accuracy, and facilitates decision-making for specialists, supporting the adoption of 40x objectives for routine laboratory use.