Project Title: Deep diagnostics: using infrared spectroscopy and machine-learning to improve malaria diagnosis in low-income communities.
Project Description: The approach that we propose combines portable infrared spectroscopy systems and machine-learning algorithms to deliver accurate pathogen detection without the use of any reagents other than simple desiccants. In 2019 we demonstrated initial performance of this approach and achieved 92% accuracy for detecting malaria parasites in dried human blood spots (DBS) collected during a survey in rural south-eastern Tanzania. The findings were verified by PCR assays. We now propose to validate this approach using samples... The approach that we propose combines portable infrared spectroscopy systems and machine-learning algorithms to deliver accurate pathogen detection without the use of any reagents other than simple desiccants. In 2019 we demonstrated initial performance of this approach and achieved 92% accuracy for detecting malaria parasites in dried human blood spots (DBS) collected during a survey in rural south-eastern Tanzania. The findings were verified by PCR assays. We now propose to validate this approach using samples collected from health facilities in areas with varying parasite prevalence rates, and using samples with different parasite densities. We will also improve our machine-learning algorithms and create a cloud based system to integrate data from different health facilities so that the machine-leaning models are gradually optimized to improve diagnostic outcomes at local level. While this project will focus primarily on malaria, the approach we are developing will likely be applicable for several other infections, possibly including neglected tropical diseases.
Principal Investigator : Issa Mshani
Department Name : HSIEP
Time frame: (2020-08-01) - (2022-12-31)
Rudolf Geigy Foundation (Normal)
External Collaborating Partners
The University Court of the University of Glasgow