Digital Image Analysis versus Visual Scoring for Bcl-2 and P53 Protein Expression in Glioma

Abstract

Background: Traditionally, evaluation of the results of immunohistochemistry was done by visual quantification. Materials and methods: for reliable evaluation, more time-efficient and user friendly method we used simple computer program with image analysis options as independent parameters for reading positive results. To test the validity of visually scored results, we compare and correlate the results of Digital image analysis (DIA) variables with the visual scores of 280 pictures taken from entire stained glioma tumor sections for Bcl-2 and P53 oncoproteins in different glioma tumor grades.
Results: In this study, rates expression of both oncoproteins was evaluated visually in glioma tumor samples (Bcl-2=72.41% and P53=82.76%), no statistical significant differences were observed according to pathologic grades. Similarly, these results were also explained by data obtained by DIA variables that has been closely correlated with visual scores. More importantly, the DIA variables explained little discrepancy in the visual scores of both oncoproteins. Conclusions: the quantitative DIA measurements of the immunohistochemically stained sections made the results more objective and supported pathomorphological diagnosis.