Particle Classification through the Analysis of the Forward Scattered Signal in Optical Tweezers

• DOI: 10.3390/s21186181
• Publication Date: 2021-09-15T16:00:44Z
• Authors: Inês Alves Carvalho; Nuno Azevedo Silva; Carla C. Rosa; Luís C. C. Coelho; Pedro A. S. Jorge
• Keywords: SIGNAL (programming language); Millisecond

The ability to select, isolate, and manipulate micron-sized particles or small clusters has made optical tweezers one of the emergent tools for modern biotechnology. In conventional setups, classification trapped specimen is usually achieved through acquired image, scattered signal, additional information such as Raman spectroscopy. this work, we propose a solution that uses temporal data signal from scattering process trapping laser, with quadrant photodetector. Our methodology rests on pre-processing strategy combines Fourier transform principal component analysis reduce dimension perform relevant feature extraction. Testing wide range standard machine learning algorithms, it shown allows achieving accuracy performances around 90%, validating concept using dynamics task. Achieved 500 millisecond signals leveraging methods low computational footprint, results presented pave way deployment alternative faster methodologies in technologies.