Automated Quantitative Analysis of HIF1-¿ Immunolabeling Using Machine Learning

• Grant number: 25/11411-9
• Support Opportunities: Scholarships abroad - Research Internship - Scientific Initiation
• Start date: September 01, 2025
• End date: October 31, 2025
• Field of knowledge: Health Sciences - Medicine - Maternal and Child Health
• Principal Investigator: Sérgio Pereira
• Grantee: Anne Beatriz de Souza
• Supervisor: Jaime dos Santos Cardoso
• Host Institution: Faculdade de Ciências (FC). Universidade Estadual Paulista (UNESP). Campus de Bauru. Bauru , SP, Brazil
• Institution abroad: Universidade do Porto (UP), Portugal
• Associated to the scholarship: 24/17187-0 - Intergenerational Effects of Sodium Saccharin on the Molecular Markers of Angiogenesis in the Rat Prostate, BP.IC

Abstract

Immunohistochemistry (IHC) is a widely used technique for detecting specific antigens in biological tissues through antibody labeling, essential for the characterization and quantification of protein expression in histological samples. However, manual assessment of IHC staining is time-consuming, subjective, and prone to errors, compromising the accuracy and reproducibility of results. In this context, computational approaches based on machine learning, particularly those employing neural networks, emerge as promising alternatives to overcome these limitations, offering greater precision, objectivity, and efficiency. This project aims to develop a computational tool for the automated detection and quantification of hypoxia-inducible factor 1-alpha (HIF1-¿) immunolabeling in digital images of prostatic tissue. Analyses will be conducted using histological samples from male Sprague-Dawley rats divided into two experimental groups: Control (C) and Saccharin (S), with the latter exposed to sodium saccharin during gestation and lactation. The tool will be applied to the immunohistochemical analysis stage of an ongoing project, enabling more accurate assessment of saccharin exposure effects on the prostate. Furthermore, the knowledge acquired on machine learning techniques may be extended to the analysis of other relevant protein markers, broadening the impact and applicability of the developed methods.