Surface-enhanced Raman spectroscopy (SERS) is a good candidate for the development of fast and easy-to-use diagnostic tools, possibly used on serum in screening tests. In this study, a potential label-free serum test based on SERS spectroscopy was developed to analyze human serum for the diagnosis of the non-small cell lung cancer (NSCLC). We firstly synthesized novel highly branched gold nanoparticles (HGNPs) at high yield through a one-step reduction of HAuCl4 with dopamine hydrochloride at 60°C. Then, HGNP substrates with good reproducibility, uniformity, and high SERS effect were fabricated by the electrostatically assisted (3-aminopropyl) triethoxysilane-(APTES-) functionalized silicon wafer surface-sedimentary self-assembly method. Using as-prepared HGNP substrates as a high-performance sensing platform, SERS spectral data of serum obtained from healthy subjects, lung adenocarcinoma patients, lung squamous carcinoma patients, and large cell lung cancer patients were collected. The difference spectra among different types of NSCLC were compared, and analysis result revealed their intrinsic difference in types and contents of nucleic acids, proteins, carbohydrates, amino acids, and lipids. SERS spectra were analyzed by principal component analysis (PCA), which was able to distinguish different types of NSCLC. Considering its time efficiency, being label-free, and sensitivity, SERS based on HGNP substrates is very promising for mass screening NSCLC and plays an important role in the detection and prevention of other diseases.
Quantum and classical machine learning for the classification of non-small-cell lung cancer patients
