Immunohistochemistry (IHC) is a widely used method for localizing and semi-quantifying proteins in tissue samples. Traditional IHC analysis often relies on manually counting 200 cells within a designated area, a time-intensive and subjective process that can compromise reproducibility and accuracy. Advances in digital scanning and bioimage analysis tools, such as the open-source software QuPath, enable semi-automated cell counting, reducing subjectivity and increasing efficiency.
This project developed a QuPath-based script and detailed guide for semi-automatic cell counting, specifically for tissues with low cellularity, such as intervertebral discs and cartilage.
The methodology was validated by demonstrating no significant differences between the manual counting and the semi-automatic quantification (p = 0.783, p = 0.386) while showing a strong correlation between methods for both collagen type II staining (r = 0.9602, p < 0.0001) and N-cadherin staining (r = 0.9044, p = 0.0001). Furthermore, a strong correlation (intraclass correlation coefficient (ICC) single raters = 0.853) between 3 individual raters with varying academic ranks and experiences in IHC analysis was shown using the semi-automatic quantification method.
The approach ensures high reproducibility and accuracy, with reduced variability between raters and laboratories. This semi-automated method is particularly suited for tissues with a high extracellular matrix to cell ratio and low cellularity. By minimizing subjectivity and evaluation time, it provides a robust alternative to manual counting, making it ideal for applications where reproducibility and standardization are critical. While the methodology was effective in low-cellularity tissues, its application in other tissue types warrants further exploration.
These findings underscore the potential of QuPath to streamline IHC analysis and enhance inter-laboratory comparability.