Recently, the integration of machine learning and nanotechnology has led to several advances in many fields including agriculture and medicine. However, the intersection of machine learning and nanotechnology has yet to breach its surface of potential. Machine learning has proven to be a useful tool for biomedical diagnostics. However, there is no artificial intelligence that can perform biomarker-free detection using experimental results. Here, we demonstrate a workflow using our algorithm termed AGONS and optical nanotechnology platform for biomarker-free detection of various biological systems including proteins, cancer cells, miRNA mimics, and more.  Previously, we had assembled a nanosensor array that achieved 70-90% predictive accuracy against these targets using simple chemometric modeling.  However, harnessing machine learning has allowed us to tremendously raise the accuracy to nearly 100% while reducing overall nanosensor array size. Utilizing AGONS provides two advantages: (1) automatic design and guiding of nanosensor array development through feature extraction techniques and (2) improved predictive accuracy. These combined advantages allow for improved downstream data collection for robust model development, sensor cost by the reduction in array size, and wider implementation space for array-based diagnostics. Through AGONS guidance in nanosensor array design, new opportunities for biomarker-free detection of the most challenging biological targets are possible.

Biomarker-free diagnostics using machine learning-guided optical nanosensor arrays

Fall 2022 abstracts, posters, and presentations will launch in mid-September.

ACS Fall 2022

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.