Automated prediction of mastitis infection patterns in dairy herds using machine learning

Area of Science:

• Veterinary Medicine
• Animal Health
• Machine Learning Applications

Background:

• Mastitis in dairy cattle incurs significant economic losses and impacts animal welfare.
• It is a primary driver of antimicrobial use in cattle farming.
• Accurate diagnosis of pathogen transmission routes (contagious vs. environmental) is crucial for effective prevention.

Purpose of the Study:

• To develop and validate a machine learning model for automated herd-level mastitis diagnosis.
• To differentiate between contagious (CONT) and environmental (ENV) mastitis transmission routes.
• To further classify environmental mastitis into dry period (EDP) or lactating period (EL) transmission.

Main Methods:

• Utilized random forest algorithms trained on data from 1000 dairy farms.
• The algorithm aimed to replicate diagnoses typically made by specialist veterinary clinicians.
• Evaluated diagnostic performance using accuracy, positive predictive value (PPV), and negative predictive value (NPV).

Main Results:

• Achieved 98% accuracy, 86% PPV, and 99% NPV for diagnosing CONT vs. ENV mastitis.
• Attained 78% accuracy, 76% PPV, and 81% NPV for distinguishing EDP vs. EL transmission.
• Demonstrated high reliability in replicating expert-level diagnoses.

Conclusions:

• An automated mastitis diagnosis tool can accurately identify transmission routes.
• This technology has the potential to assist non-specialist veterinarians in rapid diagnosis.
• Facilitates prompt implementation of targeted control measures, reducing disease impact and antimicrobial usage.