Design and implementation of an image based portable ELISA analyzer using EIPA and 4PLR

This thesis presents an implementation of predictive analytics on ELISA Imaging Systems in the absence of the standard laboratory equipment for field diagnostics. To that aim I developed a custom built optical setup with image processing and machine learning techniques. Using the light absorbance and transmittance properties of chemical compounds involved in hormone assays, I was able to estimate the hormone levels across reproductive stages. This work would allow for the eventual development of compact and economical closed systems which can be used for diagnostic advisory purposes in remote areas. This line of applied research, is expected to yield data that can be used to monitor health related outcomes. To test this use I focus the development of this tool on the monitoring of women’s ovarian function. Experimental results demonstrate that our proposed model predicts hormone levels comparable to currently used commercial and laboratory methods.