Research Area

Simultaneous detection of uric acid (UA) and ascorbic acid (AA) in different biofluids is difficult, as these biomolecules have the same oxidation potential. Utilizing Molybdenum disulfide as the sensing substrate will significantly improve the system's selectivity toward UA and AA detection. An extensive first principle analysis will promote the suitability as well as the working principle of simultaneous detection of UA and AA using Molybdenum disulfide. The electronic band gap, along with projected density of states analysis, confirms that after UA and AA adsorption, the conductivity of the system alters significantly, and thus a change in band gap is visible. Furthermore, the charge redistribution upon adsorption of two different analytes over Molybdenum disulfide substrate is validated by Löwdin population analysis. The charge difference density plot also confirms the electrostatic nature of substrate and analyte interaction.

This serves as an example of integrating a machine learning model into healthcare applications. We have created a standalone desktop application driven by machine learning. This application automates the process of glucose detection using data from oxidation current responses.

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.