Metrology for Digital Pathology. Digital pathology cross-theme project report
Michael Adeogun, J Bunch, Al Dexter, C Dondi, Tais Leal Murta, Chelsea J. Nikula, M Shaw, Adam J. Taylor, Ignacio Xavier Partarrieu, Marina Romanchikova, Neil T. Smith, Spencer A. Thomas, J Venton · 2021
Pathology underpins over 90% of healthcare decisions and is undergoing digital modernisation, increasingly relying on novel imaging and data analysis methods. The pressure on the UK pathology services has been amplified by the COVID-19 pandemic. While digitalisation of pathology services can help alleviate this pressure, many digitalisation solutions are premature, having been based on unstandardized processes, uncalibrated instruments, black-box AI tools and low-quality data. Metrology guidance on safe practices in use of Artificial Intelligence tools, industry-wide data and metadata standards, techniques to address variabilities in the image acquisition, and the training of pathologists are key to improving patient outcomes through faster and more precise, more robust diagnostics.