| 2021-RG1 | Delves, M., Dhawan, R., Gioli, M., and Whitley, E. A. (2021). Phenotype Image Data and Digital Learning Innovation (PhIDDLI), | £2,000 knowledge exchange seed funding (May 2019) and £30,000 main funding (September 2019) to support the development of an alpha version of the software.++++This project arose from a multidisciplinary research sandpit organised by the Bloomsbury Set around the use of artificial intelligence techniques to address anti-microbial resistance.
The project brought together academics from LSE, SOAS and the London School of Hygiene and Tropical Medicine to support the work of a malaria researcher.
Inspired by the experiences of the industry partner (Dhawan) the project uses existing open–source artificial intelligence software to analyse microscopic imaging of cells related to malaria, with data science expertise ensuring that the analysis is optimised for the problem space.
++In particular, the project is looking to identify the “phenotype” (i.e. the particular shape, texture, appearance) of drug–treated malaria parasites called gametocytes. These cells are responsible for transmitting the disease from the infected human to the mosquito and thus propagating malaria and allowing the spread of drug resistance genes. Research has identified many molecules that might prevent malaria transmission, the challenge is to prioritise the most promising for onward translational development.
++Microscopic imaging of cells affected by a variety of treatments/conditions/diseases generates feature–rich complex data that is laborious and time consuming to interpret manually. Automation of this process is recognised to be extremely attractive and whilst tools to develop workflows to analyse microscope images exist, all require significant technical expertise usually outside of the experience of most researchers, and few offer an “end–to–end” experience focused on the needs of the researcher.
++This project ran from 2018 to 2021
++ Resulting publication 2023-J3.++This project arose from a multidisciplinary research sandpit organised by the Bloomsbury Set around the use of artificial intelligence techniques to address anti-microbial resistance.
The project brought together academics from LSE, SOAS and the London School of Hygiene and Tropical Medicine to support the work of a malaria researcher.
Inspired by the experiences of the industry partner (Dhawan) the project uses existing open–source artificial intelligence software to analyse microscopic imaging of cells related to malaria, with data science expertise ensuring that the analysis is optimised for the problem space.
++In particular, the project is looking to identify the “phenotype” (i.e. the particular shape, texture, appearance) of drug–treated malaria parasites called gametocytes. These cells are responsible for transmitting the disease from the infected human to the mosquito and thus propagating malaria and allowing the spread of drug resistance genes. Research has identified many molecules that might prevent malaria transmission, the challenge is to prioritise the most promising for onward translational development.
++Microscopic imaging of cells affected by a variety of treatments/conditions/diseases generates feature–rich complex data that is laborious and time consuming to interpret manually. Automation of this process is recognised to be extremely attractive and whilst tools to develop workflows to analyse microscope images exist, all require significant technical expertise usually outside of the experience of most researchers, and few offer an “end–to–end” experience focused on the needs of the researcher.
++This project ran from 2018 to 2021
++ Resulting publication 2023-J3 |