K. Dao, S. Prithweeraj, B.Tech, S. Mamidi, R.R. Ampadi, H. Kanniyappan, and M.T. Mathew
UIC College of Medicine at Rockford, Illinois, United States
Poster stand number: T130
Keywords: Breast cancer, Early diagnosis, Urine based biosensor, Machine learningElectrochemical biosensors are widely utilized in various aspects of healthcare. We hypothesize developing a non-invasive protein-based electrochemical biosensor for the early prediction of breast cancer. In this study, we report the breast cancer risk levels prediction using a protein biomarker, Alkaline Phosphatase (ALP). ALP a pervasive protein present predominantly in the bone or liver is 3-fold higher than the normal in breast cancer patients. Additionally, the machine learning model can be utilized in breast cancer prediction and accurately identify the risk level. The coating of antibodies and antigens induces the change in the Redox reactions that are recorded in the form of EIS and CV to determine the change in Impedance and capacitance which differentiate the different concentrations of ALP in artificial urine. We specifically focused on changes in impedance, resistance, capacitance, and CV area values to observe the change in different concentrations of ALP in artificial urine and cloud these values into a model for machine learning. When given the EIS and CV data, the SVM was seen to have 80% accuracy in predicting breast cancer risk. Thus, this study demonstrates the potential for the use of iBiosensC in breast cancer detection.