Application of machine learning methods for classification of biological tissue using the spectra of upconversion luminescence of NaYF4:Yb3+/Tm3+ nanocomplexes
Kirill A. Buzanov1, Sergey A. Burikov1, Tatiana A. Dolenko1; 1Moscow State University, Faculty of Physics, Moscow, Russia
Abstract
Optical methods for diagnostics of the state of biological tissues based on the detection of the luminescence response of various nanostructures incorporated into biological objects are currently being actively developed. Such methods enable the detection of cancerous tissues and their mapping to determine the boundaries of neoplasms. Machine learning (ML) methods are increasingly being employed to achieve these objectives.
In recent years, lanthanide-based nanocomplexes have become increasingly widespread in biomedical applications. Their luminescence can be excited in the anti-Stokes spectral region, providing highly advantageous properties for biomedical applications: in particular, the problem of separating the luminescence signal of the nanoparticles from the tissue autofluorescence background is substantially reduced. The luminescence response of such nanocomplexes depends on the optical density of the surrounding medium, making them promising for mapping and classification of different biological tissues.
In this work, we present the results of classifying model biological tissues with different densities based on the luminescence response of NaYF4:Yb3+/Tm3+ nanocomplexes exhibiting upconversion luminescence under excitation at a wavelength of 975 nm. Beef muscle tissue and beef liver were used as model tissue samples.
The nanoparticles were injected into the biological tissues, after which luminescence maps of the nanoparticles within the tissues were acquired using a confocal microscope. Several MMO were employed for biological tissue classification, including agglomerative clustering, k-means clustering, and random forest. The obtained results demonstrate the potential of ML for the classification of biological tissues based on the luminescence response of upconversion nanoparticles.
This research was supported by the Russian Science Foundation, Grant No. 25-22-00411,
https://rscf.ru/en/project/25-22-00411/
K. Buzanov acknowledges the Foundation for the Development of Theoretical Physics and Mathematics “Basis” (Project No. 25-2-1-102-1) for support.
Speaker
Kirill Alexandrovich Buzanov
Faculty of Physics, Lomonosov Moscow State University
Russia
Discussion
Ask question
