RBGKEW1308 Expanding the use of Artificial Intelligence (AI) in root fungal colonisation assessment

RBG Kew require a consultancy service, to expand the work developed during phase 1, to adapt AMFinder, an AI visualisation tool developed by Evangelisti et al (2021) to assess fungal-colonisation, for use with more complex field root samples.

The work in phase 2 would focus on improving the accuracy of the outputs that AMFinder generates and expanding its functionality to detect different mycorrhizal types.

This will build on the preliminary work carried out in phase 1, exploring the potential of a semi-supervised learning approach, and improving the user experience of the tool, by editing the dashboard interfaces and developing pipelines.

The AMFinder tool should provide a method for rapid quantification of fungal colonisation in plant roots.

Input to the tool will be images of ink-stained fungal structures inside roots mounted on microscopy slides, taken with a digital slide scanner (P…