Development of an impedimetric immunosensor for machine learning-based detection of endometriosis: A proof of concept

Abstract A novel yet simple electrochemical immunosensor for specific and reliable detection of endometriosis is developed. Alpha-1-B glycoprotein (A1BG) exhibits higher specificity and sensitivity compared to earlier reported endometriosis serum markers. It holds promise towards clinically significant diagnosis of endometriosis. The sensing device, fabricated by covalently immobilizing monoclonal anti-A1BG antibody on a self-assembled monolayer (SAM) modified screen-printed electrode, can successfully capture the target antigen present in serum samples of endometriosis patients. The electrochemical impedance spectroscopy (EIS) method has been employed to measure the Faradic redox responses of different concentrations of A1BG. Charge transfer resistances measured from the Nyquist plots are linearly related to the logarithmic change of antigen concentration with a 1 ng/mL detection limit. The sensor requires a low sample volume of 50 μL with a resolution of 4 ng/mL. The specificity and sensitivity of the device are also tested with different interfering proteins. A robust classification model is also implemented for detection of A1BG, which could be useful in precise decision-making for end users.