Artificial Intelligence (AI) powered mobile app for the assessment of eye disease in cattle | 2021
Did you know that MLA is partnering with skilled veterinarians to develop a phone app that will allow producers on farm to diagnose the type and stage of eye disease in cattle?
| Project start date: | 03 May 2021 |
| Project end date: | 30 November 2024 |
| Publication date: | 09 August 2021 |
| Project status: | In progress |
| Livestock species: | Grain-fed Cattle, Grass-fed Cattle |
| Relevant regions: | National |
Summary
Eye disease impacts animal health and welfare and poses occupational health and safety risks for personnel working with animals that have an eye disorder. It also causes significant productivity and economic losses in Australia.
Apart from services of skilled veterinarians who can correctly diagnose eye disease, there is no other mechanism to make an assessment. This is particularly challenging in remote areas of Australia.
This project will develop a computer aided cutting edge assessment tool for eye disease, which will be delivered in the form of an app that uses artificial intelligence.
Objectives
This project is developing:
- an objective scorecard for the identification of different stages of eye diseases in cattle
- a cloud-based resource database by collecting and annotating images and video clips of eye diseases
- learning models for automatic classification of eye images into diseased and non-diseased eyes, for scoring of lesions and determination of the stage of the disease
- a smartphone app for the routine assessment of eye diseases by cattle producers and traders.
Benefits to industry
Pinkeye is the most common form of eye disease and is variable in acute versus chronic stages. Further, sometimes scar tissue can be confused as a new lesion therefore treating an animal at this stage is pointless. The app will assist with this real time understanding and assessment.
This technology will be valuable for producers, agents and other personnel that have had little or no exposure to eye disease. The tool will be developed by skilled veterinarians but won’t require vets on farm to diagnose the type or stage of eye disease, which will enable early correct diagnosis.
Cattle with visible eye lesions, whether active or not, are frequently rejected for sale or downgraded, particularly for sale to feedlots or for export. The development of an AI powered app will prevent unnecessary rejection or downgrading.
More information
| Contact email: | reports@mla.com.au |
| Primary researcher: | The University of Sydney |
