NSH Poster Podcast (2025): Histological Whole Slide Scanning Reproducibility Study

Title: Histological Whole Slide Scanning Reproducibility Study

Authors: Hannah Benton, BSa, Tomoe Shiomi, MS, HTL(ASCP)CM, CT(IAC)CM, Fatma Farooqi, BSc, HTL(ASCP), Elizabeth A. Chlipala, BS, HTL(ASCP)QIHC and Luis Chiriboga, PhD, HT(ASCP)QIHC

Abstract: Whole slide imaging (WSI) is an increasingly versatile method for capturing and sharing high-resolution digital images of stained histological slides. These images can be used for a variety of applications, including clinical diagnosis, pathology review, and image analysis. While many whole slide scanners exist with varying features tailored to different use cases, a critical factor across all platforms is the accuracy and reproducibility of the scanned images. To investigate scan consistency over time, a control slide was prepared using a tissue microarray stained with Hematoxylin and Eosin (H&E). Ink dots were applied to the slide to define a consistent scanning region. The slide was scanned 77 times over six months using an Aperio AT2 whole slide scanner at 40x magnification.Image analysis was performed using HALO software by Indica Labs. Both the entire scan area and individual tissue punches were analyzed to assess total stained area and stain intensity, quantified by optical density (OD) for both hematoxylin and eosin. A linear regression model was applied to data from all individual punches and the full scan region. Additionally, a two-way ANOVA was conducted to compare OD values of hematoxylin and eosin between the first 10 scans and the last 10 scans. Key findings were that hematoxylin showed a statistically significant decline in both stained areas and OD over time, while eosin demonstrated a statistically significant increase in stained area, but a decrease in OD. These results suggest potential degradation of staining quality or imaging consistency over time. Possible contributing factors include slide bleaching, light source variability, annotation region size, or other imaging conditions. These will be the focus of future investigations to better understand and control variability in longitudinal slide scanning studies.