Smart Detection of Mastitis in Dairy Cows: A Comparison of Traditional Methods, Smart Sensors, Thermal Imaging, and Artificial Intelligence Algorithms

Smart Detection of Mastitis in Dairy Cows: A Comparison of Traditional Methods, Smart Sensors, Thermal Imaging, and Artificial Intelligence Algorithms

Hasan Izanloo1 Ali Hossine2

IOI Code : ACZE-ACAFCA
Publication : 10th.International Congress of Developing Agriculture, Natural Resources, Environment and Tourism of Iran(10icsda.ir)
Abstract :
    Mastitis is one of the most important diseases affecting udder health, animal welfare, and the economic productivity of dairy cows. Timely detection of this disease, especially in its subclinical form, is challenging due to the absence of obvious visible signs, and reliance on a single indicator can lead to delayed identification or false alarms. This article aims to review and compare conventional mastitis detection methods with automated monitoring technologies, online sensors, thermal imaging, and machine learning algorithms. This study was conducted as a narrative review based on published scientific literature concerning the pathophysiology of the disease, diagnostic indicators, automatic milking systems, milk quality sensors, infrared thermography, and data analysis. Evidence indicates that tests such as somatic cell count (SCC), the California Mastitis Test (CMT), microbial culture, and clinical evaluation remain the foundation of veterinary diagnosis and decision-making; in contrast, sensors and artificial intelligence models can enhance continuous monitoring and early detection of changes. However, the performance of automated systems depends on sensor type, case definition, herd prevalence, data quality, and alert thresholds. Moreover, udder surface temperature or electrical conductivity changes alone do not have sufficient specificity to determine the cause of the disease. The appropriate approach is the integration of multi-source data with veterinary examination and confirmatory tests. External validation in herds and under different management conditions, reporting of performance metrics beyond overall accuracy, and cost-benefit evaluation are essential for the practical development of these technologies.

Keywords : bovine mastitis somatic cell count smart sensors thermal imaging machine learning