Detection and classification of nanoparticles internalized in tumor cells using computational methods

• Grant number: 24/01849-4
• Support Opportunities: Research Grants - eScience and Data Science Program - Regular Program Grants
• Start date: July 01, 2024
• End date: September 30, 2025
• Field of knowledge: Biological Sciences - Biophysics - Biophysics of Processes and Systems
• Principal Investigator: Allan Felipe Fattori Alves
• Grantee: Allan Felipe Fattori Alves
• Host Institution: Instituto de Biociências (IBB). Universidade Estadual Paulista (UNESP). Campus de Botucatu. Botucatu , SP, Brazil
• Associated researchers: Guilherme Augusto Soares ; Rachid Jennane

Abstract

Regarding in vivo applications, nanoparticles can be used for the diagnosis and treatment of various cancer types, providing the so-called theranostic property. Transmission electron microscopy (TEM) is a powerful technique for observing nanoscale samples, particularly cellular structures. The literature has brought significant challenges regarding determining the internalization of nanoparticles in tumor cell cultures, which is a costly process and highly dependent on specialized professionals. In this context, the automatic analysis of electron microscopy images with the aid of computational methods has demonstrated enormous benefits for research into the characterization of these techniques. The main objective of this project is to develop a segmentation and classification algorithm based on deep learning architectures that will be applied to databases containing transmission electron microscopy images. These databases will be composed of samples of tumor cells with nanoparticle endocytosis processes. This project includes five stages of development that encompass the synthesis of nanoparticles, their administration in tumor cell cultures, the organization of microscopy image banks, the pre-processing of these banks aiming to reduce noise and highlight structures, and finally, the development of algorithms with Deep Learning techniques for the segmentation and classification of nanoparticle endocytosis processes in an automated manner. These results would provide excellent reliability in interpreting images and make analyses that can be tedious and subject to interpretation errors faster and more effective. (AU)