CONGRESS 2023 - The need for end to end QC in digital histopathology and artificial intelligence (AI)

Histopathology has numerous stages in the production of a digital image and its subsequent use. Each of the stages can introduce variations that are compounded resulting in a net variation in image quality for nominally the same tissue. Humans are tolerant of variation so this variation in quality has minimal impact on outcomes, which are additionally validated by EQA services. But AI is in some cases being negatively impacted by variation and highlights the need for quality metrics and subsequently standards for each stage, where possible. But currently there are few independent QC tools for digital histopathology. This presentation will present the results of our work in NPIC were we have developed QC tools for staining, digitisation and display in digital histopathology.

Download as vCalendar (for Microsoft Outlook etc.)

Import event to your Google Calendar