AI-Supported Ultrasound Analysis for Automated Pregnancy Detection in Cows

Initial Situation and Objectives

Pregnancy diagnosis using ultrasound is a key component of fertility management in dairy farming. Reliable examinations require solid expertise and practical experience. At the same time, the availability of specialized professionals is decreasing while the demand for regular examinations continues to increase.

The aim of the project is to develop an AI-supported ultrasound analysis system that assists in automating pregnancy detection and ensures consistent diagnostic quality.

Technological Approach

The project develops AI models that automatically analyze ultrasound images and videos. The system will:

• detect fluid-filled areas,
• differentiate between fluids and image artifacts,
• segment and measure relevant structures of the uterus.

The analysis is performed directly during the scanning process. The models are specifically trained for agricultural applications and designed for integration into existing ultrasound devices.

_{Detection of the Fetus in Ultrasound Using A}

Benefits, Application, and Exploitation

AI-supported analysis reduces training requirements and enables faster onboarding of users. Diagnoses can be carried out more efficiently, and fertility monitoring can be scheduled more frequently and systematically. Depending on the application concept, the technology can also be used directly on farms as a supportive tool.

_{Automated Ultrasound Analysis in Cows}

Practical Environment

The development is carried out in collaboration with Fraunhofer IGD, Visual Healthcare Technologies division.

Funding: