Quality control for digital pathology

Presenter: Aleksandra Zec
DEIB PHD Student
DEIB - BIO1 Room (Bld. 21)
March 20th, 2025 | 11.00 am
DEIB PHD Student
DEIB - BIO1 Room (Bld. 21)
March 20th, 2025 | 11.00 am
Sommario
On March 20th, 2025 at 11.00 am Aleksandra Zec, PHD Student in Information Technology, will hold a seminar on "Quality control for digital pathology" at DEIB BIO1 Room (Building 21).
The field of digital pathology, which relies on the use of Whole Slide Images (WSIs), plays a crucial role in advancing healthcare. WSIs are digitized biopsy samples stored in a pyramid format, containing varying magnifications and resolutions that capture different levels of information. These images represent a rich and valuable source for predicting critical patient outcomes, such as therapy response and survival time. However, the manual preparation of these slides, despite standardized protocols, introduces variability across institutions, as well as potential artifacts. The motivation for implementing WSI quality control is to identify low-quality WSIs or problematic areas within them, ensuring that differences in preparation protocols do not interfere with analyzing the actual appearance and biological content of the image, ultimately standardizing the quality of data coming from different institutions.
Quality control issues arise at various stages, including tissue preparation, staining, and scanning. This seminar will explore computational techniques to address these challenges, focusing on stain normalization, color artifact removal, tissue folding correction, and handling out-of-focus areas, ultimately enhancing the quality of WSIs for accurate analysis.
The field of digital pathology, which relies on the use of Whole Slide Images (WSIs), plays a crucial role in advancing healthcare. WSIs are digitized biopsy samples stored in a pyramid format, containing varying magnifications and resolutions that capture different levels of information. These images represent a rich and valuable source for predicting critical patient outcomes, such as therapy response and survival time. However, the manual preparation of these slides, despite standardized protocols, introduces variability across institutions, as well as potential artifacts. The motivation for implementing WSI quality control is to identify low-quality WSIs or problematic areas within them, ensuring that differences in preparation protocols do not interfere with analyzing the actual appearance and biological content of the image, ultimately standardizing the quality of data coming from different institutions.
Quality control issues arise at various stages, including tissue preparation, staining, and scanning. This seminar will explore computational techniques to address these challenges, focusing on stain normalization, color artifact removal, tissue folding correction, and handling out-of-focus areas, ultimately enhancing the quality of WSIs for accurate analysis.