Advanced Data Analytics and Organic Color Centers for Diagnostic Applications

Abstract

Biofluids, most notably serum, contain many molecules that together may help to define the physiological state of a person. A grand challenge is to use serum to identify a clinical state with a high degree of certainty. Rather than attempting to accomplish biomarker development on a one-to-one recognition basis, it may be more effective to determine a whole clinical state via perception-based platforms due to recent achievements in sensor technologies and computational algorithms. In this talk, we present the development of carbon nanotube-based nanosensor arrays wherein the optical responses of the nanotubes were used to train machine learning models. We introduce the optical and chemical diversity of nanosensor array using molecular masking effects of DNA-wrapping and covalent functionalization of fluorescent quantum defects on nanotube surface. Machine learning can be employed to process the complex spectral responses of nanosensors and recognize spectroscopic fingerprints of certain clinical states from serum. We will discuss how these methods can complement and improve conventional diagnostic assays.