Machine Learning Approach for Early Lactation Mastitis Diagnosis by Applying Total and Differential Somatic Cell Counts

• DOI: 10.20944/preprints202503.1970.v1
• Publication Date: 2025-03-28T00:52:39Z
• Authors: Alfonso Zecconi; Francesca Zaghen; Gabriele Meroni; Flavio Sommariva; Silvio Ferrari; Valerio Sora
• Keywords: Somatic Cell Count

Dairy herds around the world are undergoing several changes. Herd sizes increasing, as both milk yield and quality. The implementation of new technologies in various domains dairy production is leading to an increase amount data available. This, turn, creates a need extract useful information from these improve efficiency. This paper presents findings preliminary study that utilizes machine learning (ML) approach assess accuracy somatic cell count (SCC) neutrophils + lymphocytes Count/mL (NLCC) identifying cows at risk developing intramammary infection (IMI) due major pathogens (S. These include S. aureus, agalactiae, uberis, dysgalactiae. identified either by real-time PCR (qPCR) methods or conventional bacteriology, following cows' calving process. encompassed total 424 1,696 quarter samples. A comparison two revealed significant disparities prevalence MajP, with qPCR method demonstrating higher than bacteriology. However, negative results was comparable, yielding approximately 71.0% 72.1%, respectively. comprehensive substantiated all cellular markers exhibited most precise values when MajP IMI diagnosed using nearly equivalent performance, irrespective ML algorithm employed. indicate approaches based on SCC PLCC may be for healthy quarters. it essential confirm "non-negative" through subsequent analysis within 7-15 days ensure accuracy. further studies necessary enhance diagnostic