Livestock sector development and priority setting is heavily dependent on a good understanding of current animal farming and livestock production practices and information systems available and in development. This blog post sets out to describe and include tools to monitor, gather, transform and above all share vital information between key stakeholders that can help the sector to become more sustainable.
From agricultural development that informs livestock feed to vaccine treatment, the current state of animal farming is outlined with the identification of several ongoing challenges and opportunities for the future to be met.
Pre-slaughter intervention innovation is on the rise. Promising research has focused developing modified feeds for additional dietary benefits, immunology, and longevity of companion animals in emerging economies against disease-causing bacteria among food animals (Int J Mol Sci., 2015).
Feed additives to enhance animal production, tenderness, color, carcass leanness, and water holding capacity are also being developed with the side benefit of increasing ratio of protein to fat in meat along with other nutrients to enhance shelf-life, nutritional profile, taste, and visual appearance (E. Pospiech and M. Montowska, 2011).
Food safety systems, technology, and monitoring tools
Biosensors integration within the food supply chain to monitor for changes to safety of food and animal resources, products, and sharing of food is defining leading practices (Mishra et al, 2018). These integrations include rapid and efficient transcriptomics, proteomics, and biosensor-based technologies for detecting a wide variety of contaminants.
Bioinformatics systems and solutions
Bioinformatics systems based on DNA microarrays containing annotated sequences, marker vaccines, and RNA gene manipulation mappings to prevent and treat disease are being integrating within information systems between industry, academic, and government research resources.
Supply chain integration – from testing to containment with regulatory agencies to accelerate identification and containment of infected animals is a key component to these bioinformatics systems and solutions.
Protein manipulation at the molecular level
Finally, the development of effective pharmaceutical products based on antibiotics and probiotics, plus the investigation into breeding plants specifically to deliver additional nutrients to livestock is improving feed structures by utilizing nutraceuticals designed to address specific health and wellness concerns for an animal.
Challenges and opportunities
The aim of digital pathology (an image-based information environment which is enabled by computer technology that allows for the management of information generated from a digital slide) is to investigate image analysis, with the aid of specialized algorithms developed under the auspices of machine learning (ML) to develop a multi-modal platform for advanced biomedical image measurement and pattern analysis system for improving cancer diagnosis and translational applications in animals and humans. At the University of Surrey, a digital pathology project led by Prof Roberto La Ragione and Dr Kevin Wells is building on machine-led analysis developed (Alnowami et al, 2018,. Bober et al 2017) to exploit information present in images, and other information, used for the diagnosis and characterisation of particular tumour types using digital pathology data sets available at the University and through their veterinary partners.
Using artificial intelligence (A.I.) for veterinary surveillance
In recent years there has been great interest in the use of unbiased next-generation sequencing (NGS) technology for comprehensive detection of pathogens from clinical, food and environmental samples (Dunne et al., 2012; Wylie et al., 2012; Chiu 2013; Firth and Lipkin, 2013). Conventional diagnostic testing for pathogens is narrow in scope and fails to detect the aetiologic agent in a significant percentage of cases (Barnes et al.,1998; Louie et al., 2005; van Gageldonk-Lafeber et al., 2005; Bloch and Glaser, 2007; Denno et al., 2012).
Failure to accurately diagnose and treat infections in a timely fashion contributes to continued transmission and increased mortality in hospitalized patients (Kollef et al., 2008). Thus, rapid accurate and accessible diagnostics are urgently required. For rapid analysis of clinical samples collected during routine surveillance post-mortem examinations, imaging and genetic data can be considered as structured data. Physical examination, clinical laboratory reports, operative notes and discharge summaries can be considered as unstructured data.
The use of artificial intelligence with structured and unstructured data can be in three ways. Firstly, Machine Learning (ML) methods analyse structured data to cluster them to find probabilities of a certain outcome based on structured data. Secondly, information from unstructured data can be extracted using natural language processing (NLP) methods to supplement and enrich structured data. Thirdly, the most recent deep learning is used to handle more complex and non-linear patterns in data. Supervised learning used for feature extraction and unsupervised learning used for predictions are the two main divisions in ML algorithms. This proposal aims to develop AI methodologies for the rapid analysis of clinical samples collected during routine surveillance post-mortem examinations.
Monitoring AMR on dairy farms
Sensory data from dairy cows and other periodic data on AMR management (prescription, dosages, administration etc.) will be processed in a machine learning based management cloud. A low cost sensor system will be deployed in a number of automated dairy farms to gather information about health and behaviour of individually identified cows in order to understand their pathogen exposure, antibiotic treatment and incidence of AMR. Farms with automated food supply and milking system will also be used to provide milk production data to the management cloud.