Live Cell Breast Cancer Phenotyping Using Spectral Fingerprinting of SWCNTs Coupled with Machine Learning

Document Type

Presentation

Date of Original Version

3-27-2026

Abstract

Detecting Breast cancer at an early stage is a significant challenge because of its multiple subtypes and elusive heterogeneous nature. An innovative single-walled carbon nanotube (SWCNT) NIR fluorescence spectral fingerprinting approach in conjunction with a machine learning algorithm could precisely detect breast cancer subtypes in vitro. We demonstrate this concept by incubating DNA-functionalized SWCNTs with a panel of human breast cell lines: the non-tumorigenic MCF-10A and cancer cell lines MCF-7, HCC1954 (HER2+), MDAMB-231, and MDA-MB-468 (triple-negative). The NIR fluorescence spectra of SWCNTs across 500 individual cells of the mentioned cell types showed significant differences in emission peak intensities, center wavelengths, and peak intensity ratios, attributable to variations in cellular uptake and biomolecular interactions. These features were used to train a support vector machine learning (SVM) model. The model achieved more than 90% classification accuracy, selectivity, and precision, effectively distinguishing cancerous subtypes from noncancerous cells. Insights from this research enhance the development of nanomaterial-based platforms for biosensing and provide potential for real-time monitoring of in vivo cellular differentiation.

This document is currently not available here.

Share

COinS