Recent Breakthroughs - AMF & AI-Assisted Root Staining

Part of the reason why AMF uptake has been slow is partly down to the method it is assessed. The fact that a high level of expertise and facilities have been required in order to quantify how inoculated your crops' roots' are with AMF meant that trials designed around limited feedback response times made progress slow and an uphill battle Prior to now, the process to assess the success of root colonization by AMF was a long, arduous and expensive process involving staining the roots, doing a visual assessment and manually counting with a clicker each time the hyphae had penetrated the cortical root cells in a certain area, unsurprisingly giving a large margin for human error and taking a long time to calculate a fairly basic question - What percentage of your roots are inoculated with AMF? However, those days may soon be behind us due to the latest breakthrough in research out of Sainsbury's Laboratories and Cambridge University. They have recently trained an AI to do the visual assessment automatically by training it on a plethora of images of inoculated roots. The assessment merely requires staining the roots and taking a high resolution of said stained roots under a high-powered microscope for the AI to give a result of root inoculation success, which already will save countless hours of time and money. This new feedback mechanism could really cheapen the cost of these trials and allow more design around factors affecting colonization success with more accurate response data, in field conditions.

Full text avlaible below:

https://nph.onlinelibrary.wiley.com/doi/epdf/10.1111/nph.17697