Abstract
Immunohistochemical evaluation is essential for the diagnosis and prognostic assessment of breast carcinoma; however, conventional assessment relies on subjective estimation of staining intensity and the proportion of positive cells, which may not fully capture spatial patterns within the immunohistochemical signal. This study aimed to explore the utility of Hessian-based morphological analysis as a quantitative approach to complement conventional IHC evaluation. A total of forty breast carcinoma cases in women immunostained for p53 and BCL-2 were included. Digitized slides were analyzed using FIJI (ImageJ) to extract the DAB channel and quantify both intensity-based and Hessian-derived morphological features. Statistical analysis was performed using the Kruskal-Wallis test followed by Dunn's post-hoc test with Holm correction, and effect size was estimated using epsilon-squared. Hessian-based features demonstrated significant differences among p53 expression patterns, with higher values observed in cases with p53 overexpression compared to wild-type (p < 0.01, ε² ≈ 0.21). The largest difference was observed between the wild-type and overexpression groups, while the null group showed substantial overlap with both groups. A significant pairwise difference was observed between the wild-type and overexpression groups, whereas the wild-type-null and null-overexpression comparisons were not statistically significant. In contrast, no significant difference was observed among BCL-2 expression groups. These findings suggest that Hessian-based morphological analysis may serve as a quantitative complementary method to conventional IHC assessment, in this exploratory cohort, by capturing DAB-derived local image-structure features that are not represented solely by conventional intensity-based assessment. Further studies with larger independent cohorts, predefined image-analysis procedures, molecular validation, and reproducibility assessment are required before the clinical relevance of this approach can be determined.