FTIR and RAMAN optical spectroscopy associated with artificial intelligence algorithms in the identification of the reproductive status of bovine females in different biological samples

• Grant number: 25/13769-8
• Support Opportunities: Regular Research Grants
• Start date: October 01, 2025
• End date: September 30, 2026
• Field of knowledge: Agronomical Sciences - Veterinary Medicine - Animal Reproduction
• Principal Investigator: Luciano Andrade Silva
• Grantee: Luciano Andrade Silva
• Host Institution: Faculdade de Zootecnia e Engenharia de Alimentos (FZEA). Universidade de São Paulo (USP). Pirassununga , SP, Brazil
• Associated research grant: 24/16812-9 - OPTICAL SPECTROSCOPY RAMAN ASSOCIATED WITH ARTIFICIAL INTELLIGENCE ALGORITHMS TO PREDICT CONCEPTION AND GESTATIONAL STATUS IN BOVINE FEMALES, AP.R

Abstract

Global food demand grows alongside population growth, driving the livestock industry to adopt technologies that increase productivity and efficiency. In beef cattle farming, advances such as the multiplication of genetic material and the use of reproductive biotechnologies have been essential. However, the effectiveness of these techniques can still be improved. One way to improve bovine reproduction is through the early and accurate identification of the phases of the estrous cycle, which directly influence reproductive management and conception rates. Rapid, accurate, and viable diagnostic methods for field use are essential in this context. With the advancement of computer systems, artificial intelligence and machine learning are emerging as powerful tools for analyzing complex data, such as spectra obtained through optical spectroscopy (a fingerprint) of biological samples. Applying these techniques to identify the phases of the estrous cycle can significantly contribute to future studies and the development of more efficient reproductive diagnostic methods. This approach aims not only to improve the accuracy of field diagnoses, but also to open doors for innovations in research outside of bovine reproductive management. In this study, spectrographic patterns obtained by FTIR and RAMAN optical spectroscopy using biological samples of urine, serum, blood plasma, saliva, and vaginal swabs from female cattle at different reproductive stages will be compared. Ten multiparous cows will be selected, from which samples will be collected during estrus and diestrus, ten prepubertal anestrous heifers, and ten pregnant cows. The objective is to determine which biological sample and which spectroscopic technique presents the best accuracy in detecting the different reproductive stages. The central hypothesis of this project is that different reproductive stages determine unique metabolic patterns in female cattle that can be identified by optical spectroscopy coupled with analysis using artificial intelligence algorithms. The obtained spectra will undergo preprocessing, such as noise removal, normalization, and smoothing, to ensure greater analysis accuracy. For statistical analysis, principal component analysis will be applied to reduce dimensionality and identify relevant patterns. Machine learning algorithms can then be used to classify the samples according to the phases of the estrous cycle in order to identify the one that presents the greatest precision and accuracy in this classification. (AU)