Descrição
Optical biosensors with plasmonic substrates are typically used with detection methods based on spectral measurements with spectrometers and sophisticated optical instruments. To allow this type of biosensors in POC applications it is necessary to employ simple optical detection methods and data analysis. The use of optical microscopes as tools for measuring the response of the AuNI/glass biosensor devices can offer a simple alternative, as the images can be analyzed using machine learning (ML) algorithms for image classification. Here, we propose a system exploiting computer vision with ML, including Deep Learning algorithms, for microscopy image classification as measurement method for plasmonic genosensors. We demonstrate the use of ML for optical microscopy image classification of plasmonic substrates formed by AuNI/glass after surface modifications in a genosensor for SARS-CoV-2 virus. (1) The optical transmission microscopy images (40x magnification) were classified with Deep Learning and other ML algorithms. (2) The networks were able to distinguish the images of the genosenosors after the interaction with SARS-CoV-2 ssDNA from those which interacted with non-complementary ssDNA sequences or other three control groups (clean, 11-MUA, Blank). The accuracy was 0.87 with up to 96.8% sensitivity. In summary, simple optical microscopy coupled with a deep neural and ML networks can be employed in detection using plasmonic genosensors. Further studies are required to help identify the optimal classification descriptors for plasmonic sensors and other biosensing applications.
Referências
1 SOARES, J. C. et al. Detection of a SARS-CoV-2 sequence with genosensors using data analysis based on information visualization and machine learning techniques. Materials Chemistry Frontiers, v. 5, n. 15, p. 5658-5670, 2021.
2 RIBAS, L. C.; SÁ JUNIOR, J. J. M.; SCABINI, L. F. S.; BRUNO, O. M. Fusion of complex networks and randomized neural networks for texture analysis. Pattern Recognition, v. 103, p. 107189-1-107189-10, July 2020.
| Certifico que os nomes citados como autor e coautor estão cientes de suas nomeações. | Sim |
|---|---|
| Palavras-chave | Biosensors. Plasmonics. Machine learning. |
| Orientador e coorientador | Osvaldo Novais de Oliveira junior |
| Subárea 1 | Física da Matéria Condensada |
| Subárea 2 (opcional) | Física Aplicada à Biologia e à Medicina |
| Subárea 3 (opcional) | Física Computacional |
| Subárea 4 (opcional) | Instrumentação e Detectores |
| Agência de Fomento | CAPES |
| Número de Processo | "Não se aplica" |
| Modalidade | DOUTORADO |
| Concessão de Direitos Autorais | Sim |
