Dr Philip H. Jones, Animal and Plant Health Agency (APHA.gov.uk)

One Health and Planetary Health: the case for considering a unified health perspective

There is always a tendency for healthcare to be independently focused and organised on specific sectors, whether human, veterinary or environmental health. However, there are many examples where advances made in one field have had direct benefits to other sectors. The application of new technologies is no different. This talk makes the case for considering a unified health perspective to deliver maximum benefits to all.

Phil Jones graduated as a veterinary surgeon from the University of Bristol in 1991. After several years in mixed veterinary practice, he returned to Bristol to complete a PhD in veterinary science followed by a masters degree in preventive veterinary medicine at the University of California, Davis. Phil was appointed lecturer in Veterinary Epidemiology and Public Health at the University of Liverpool in 2006 where he worked closely with the SAVSNET team. In 2018, Phil moved to his current position in the Animal and Plant Health Agency as lead epidemiologist for scanning surveillance. He leads the Surveillance Epidemiology and Data Analysis team which manages APHA's routine scanning surveillance outputs but also investigates opportunities to develop and deliver added surveillance value from new and existing data sources.

Dr Susan Armstrong (Veterinary Health Innovation Engine (vHive), University of Surrey)

Translational Diagnostics in Companion Animals

Translational diagnostics is the ability to take basic research from lab bench to a clinical application. Currently many traditional blood biomarker tests for human and animals diseases have poor sensitivity and specificity or the tests required for definitive diagnosis are costly or invasive. By taking a multi-omic biomarker research approach, we can look to find better ways of diagnosing diseases earlier in our veterinary (and human) patients with more accuracy. Excitingly, these improved biomarkers may even allow for prognostication and better stratification of patients to enhance their care. MicroRNAs (miRs) are emerging as such circulating biomarkers for a wide range of diseases in both human and veterinary patients. Their abundance and stability in all biological fluids, in conjunction with their relative organ and/or disease specificity, alongside conservation across species makes them excellent biomarker candidates. Exemplars discussed will be miR-122 for the diagnosis of hepatopathies in dogs, cats and horses which was translated into a PCR based assay as part of a not-for-profit diagnostic lab (www.mirvetlab.com). I will also describe the pipeline of microRNA biomarker research being conducting on pancreatitis and chronic kidney disease in dogs and how the translation of these to clinical tests is a reality with the development of in-house assays and point-of-care monitors. By embracing comparative translational diagnostics and a multi-omics approach with model-based data integration, we are uniquely positioned in the veterinary world to make significant impact on addressing unmet clinical needs and improving the standard of clinical care for animals (and potentially even humans) we will never meet and how the potential of translational precision medicine is a reality.

A clinical researcher and equine veterinarian, Susan Armstrong has developed a strong research portfolio in biomarkers and translational diagnostics in companion animals with emphasis on proteotranscriptomics and point-of-care assays. Embracing the one medicine philosophy, Susan conducts research with clinical relevance in liver, lung, pancreatic, ocular and renal disease, all with human medicine collaborators. She was an integral part of the non-for-profit mirvetlab.com biomarker lab set-up at the University of Edinburgh where she completed her PhD on CKD biomarkers. Susan is now a Senior Lecturer in Veterinary Clinical Research at the University of Surrey (UoS) and Veterinary Director of the Veterinary Health Innovation Engine - a collaboration between UoS and Zoetis. Among other roles she is also a council member of the Royal Society of Medicine - Comparative Medicine Section. Her current principle research emphasis is on biomarkers of canine CKD and exercise-induced pulmonary haemorrhage in horses.

Dr Mercedes Arguello Casteleiro (BCS SGAI) and Tim Furmston (University of Manchester)

Hands-on Tutorial: Low-Code/No-Code AI: Democratise AI for Text, Images, and Audio

Low-code/no-code Artificial Intelligence (AI) is on the rise and Large Language Models (LLMs) are at its core, including multimodal LLMs that can produce content (text, image, video, or audio/speech) as output (generative AI). Emerging approaches for low-code/no-code AI typically employ visual interfaces (e.g. drag-and-drop), instead the BCS SGAI (the Specialist Group on AI of British Computer Society) is supporting an alternative that advocates for few lines of python code, using open-source LLMs and python libraries, that can be written and executed in a Web browser with Google Colab (free of charge for basic use and without setup to use). Updating or altering a few lines of python code can be as effective as enacting changes in a drag-and-drop interface.

Come along if you are interested in the latest state-of-the-art for Natural Language Processing or Computer Vision or Machine Translation (e.g. text-to-speech) using open-source Multimodal LLMs, and finding out about content generation and content analysis in 3 lines of python code. Some of the exercises use open-source (free-of-charge) LLMs from Google, Facebook, Microsoft and OpenAI. The 3 lines of python code can also be executed in proprietary machines, such as Manchester Data Science servers.

Dr Mercedes Arguello Casteleiro has a PhD in Physics and is an elected committee member of BCS SGAI (the Specialist Group on Artificial Intelligence of the British Computer Society). Lecturer (2022-2024) in Electronics and Computer Science at the University of Southampton. Before 2022, she was a researcher in the Bio-Health Informatics Group in the School of Computer Science at the University of Manchester. She is interested in "One Health" and "Planetary Health", investigating more transparent and explainable AI models by integrating deep learning with symbolic reasoning (Neurosymbolic AI).

Dr PJ Noble (Small Animal Veterinary Surveillance Network (SAVSNET), University of Liverpool)

Mixed AI approaches to surfacing information hidden in veterinary electronic health records

The Small Animal Veterinary Surveillance Network (SAVSNET) is a pioneering initiative hosted at Liverpool University that collates electronic health records (EHRs) from approximately 250 veterinary practices across the UK, encompassing data from nearly 2 million companion animals and over 12 million clinical records. As the volume and complexity of this data grows, the application of artificial intelligence (AI) methodologies has become central to our mission of extracting meaningful insights to support disease surveillance, animal health and epidemiological studies.

I will share SAVSNET’s experience using AI techniques such as topic modelling and domain-adapted (fine-tuned) language models to analyse unstructured clinical free text, particularly aiming to maintain explainability. These models enable us to identify trends and patterns across large datasets, improving our understanding of veterinary clinical practice, disease prevalence, and health outcomes in companion animals.

Building on these methods, SAVSNET is also leaning on generative large language models (gLLMs) to annotate and extract key information from free-text records with greater precision and efficiency.

This talk will explore the varied neural language-model based approaches we have deployed to date and the potential of combining traditional AI models with advanced gLLMs to improve data-driven insights from veterinary EHRs. I will also touch on how these annotate data and tools might feed into veterinary education.

Dr. PJ Noble is a Senior Lecturer in Small Animal Internal Medicine at the University of Liverpool Small Animal Teaching Hospital, with 30 years of experience in referral clinical practice. In addition to his clinical work, P-J has been a key driver in the development of the Small Animal Veterinary Surveillance Network (SAVSNET), which collects data from electronic health records across the UK. Leveraging his programming skills, P-J has written software for data presentation and annotation, supporting undergraduate and postgraduate research projects. Through collaborations with Manchester and Durham universities, he has also led the development of neural language models to extract large-scale insights from the SAVSNET database.

Peter Jones (NHS Professionals, University of Bolton)

AI for One Health and Planetary Health: is there a generic framework?

The 21st century brings major global challenges, e.g. rapid population ageing and climate change are occurring together, calling for transdisciplinary involvement of diverse sectors and stakeholders, including academics and students, companies and societal organisations. One Health and Planetary Health are "are highly complementary fields of scientific inquiry with solid leverage for translation into policy and practice".

This talk proposes the Hodges' Health Career Model (a.k.a. Care Domains Model) as a generic framework applicable to AI for One Health and Planetary Health. Hodges' Health Career Model distinguishes four domains relatable to digital healthcare and well-being data.

Despite the fragmentation and disparity of digital healthcare and well-being data, it can be used as specialised training data for Large Language Models (LLMs). This talk will include a low-code/no-code AI live demonstration of text classification with 3 lines of python code using different LLMs.

Peter Jones (mental health and general nurse trained) works as a community mental health nurse, with 47 years experience in the NHS. Supervised by the University of Manchester, his B.A. (Hons.) dissertation in computing and philosophy, was entitled: "Using a Semantic Network to Represent Nursing Terminology". A combined interest in care processes and technology enhanced learning since the early 1980s is ongoing. As a practicing nurse, Peter adopts a psychosocial and holistic approach to person-centred care and utilises Hodges' model of care - a generic conceptual framework. Hodges' model, facilitates reflective practice and critical thinking with support for lifelong learning. Drawing on clinical experience and expertise, a theoretical basis for Hodges' model is sought. A web presence, and small bibliography are encouraging other researchers to discover and apply Hodges' model in their own respective disciplines. Peter believes Hodges' model, can help address the challenges of integrated care, information disorder, parity of esteem, the determinants of health, health literacy, global and planetary health, and the need for sustainable health systems. The new academic year brings the exciting and challenging prospect of part-time tutoring at the University of Bolton.

Dr Manoj Kulshrestha (Consultant Ophthalmologist, NHS Professionals, University of Buckingham)

Low-Code/No-Code AI for Biomedical Image Classification: Visual Impairment as a case study

Over the decades, the quality of biomedical images has improved significantly. Image recognition has been a focus of AI for many decades, and biomedical image classification is a task that can benefit for the latest AI advancements, e.g. multimodal Large Language Models (LLMs).

According to the World Health Organization (WHO): "globally, at least 2.2 billion people have a near or distance vision impairment. In at least 1 billion of these, vision impairment could have been prevented or is yet to be addressed". WHO estimated a global annual cost of $411 billion (USD) for visual impairment. This talk will look into biomedical datasets of 2D images (standard benchmarks) with the aim of investigating if AI can assist the human by exhibiting a consistently demonstrated high performance in biomedical image classification with some benchmarks. For example, RetinaMNIST is a challenging dataset of 1,600 fundus camera images for the ordinal regression task. Ordinal regression is a common task in medicine, grading some image on an ordinal scale (e.g. from 0 to 4, where 0 is the absence of a disease). The fundus camera (fundus photography) can aid the detection and screening of various causes of treatable and preventable blindness, such as: diabetic retinopathy, age-related macular degeneration, glaucoma, and retinopathy of prematurity.

LLMs may bring human and AI together to gain more valuable insights than either could alone. This talk will include a low-code/no-code AI live demonstration of image classification with 3 lines of python code using different LLMs.

Mr Manoj Kulshrestha is an experienced consultant ophthalmologist with over 30 years experience. His expertise covers various areas of expertise, including medical retinal disease including age related macular degeneration, uveitis, retinal vein occlusion and diabetic retinopathy.

Mr Kulshrestha is a fellow of both the Royal College of Ophthalmologists (FRCOpth), the Royal College of Surgeons of Ireland (FRCSI). He also holds a Bachelor of Medicine and Bachelor of Surgery (MB ChB) from the University of Birmingham.

Mr Kulshrestha has actively contributed to medical education as deputy undergraduate tutor at Leighton Hospital. He also previously lectured in ophthalmology at the Bristol Eye Hospital.

His research contributions are notable with over 85 publications in peer-reviewed journals, book chapters, and newspaper articles. He has delivered lectures at various levels, including international training sessions for optometrists in Oslo, Copenhagen, Stockholm, and Helsinki. Additionally, Mr. Kulshrestha has participated as a co-investigator in research studies concerning intravitreal medications and dry eye disease.

In his career, Mr. Kulshrestha has held significant roles, such as Regional Director and Chair of Ophthalmology, Lead Clinician and Population Health Chair for Head, Neck, Eye, and Skin for the Hywel Dda University Health Board. His commitment to teaching, research, and patient care reflects his dedication to excellence in the field of ophthalmology. He is currently the Clinical Lead and Quality Improvement Lead for Ophthalmology at Leighton Hospital.

Dr Saihong Li, University of Stirling

Low-Code/No-Code AI for Planetary Health: World Fisheries and Aquaculture as a case study

In 2006, Google launched the Google Translate — a free-to-use machine translation website. Google used AI to transform Google Translate. In the last years, the world has witnessed significant advancements in AI with the development of large language models (LLMs) trained on text, images, audio, videos, and programming code. Generative AI with OpenAI’s ChatGPT is well known to the public. The latest LLMs created are multimodal (e.g. Google’s Gemini 1.5 Pro) are even more versatile.

In the UK, the language services market rose from the estimated GBP 1.7 billion in 2021 to a current estimate between GBP 1.94 and 2.20 billion. This talk focuses on LLMs and its impact on Translation Studies considering the adequacy, i.e. the quality of the machine translation output. To gain some valuable insights, we used a flagship publication of the Food and Agriculture Organization of the United Nations about fisheries and aquaculture, i.e. a sector that is contribution to fighting poverty, hunger and malnutrition. According to flagship publication: "today, 811 million people suffer from hunger and 3 billion cannot afford healthy diets".

The case study utilised low-code/no-code AI, and the talk will include a live demonstration of machine translation with 3 lines of python code using different LLMs.

Dr Saihong Li is a Senior Lecturer in Translation at the University of Stirling. Her research interests include interdisciplinary digital humanities, terminology and cross-cultural and translation studies. Dr Li has produced a substantial body of research related to terminology and translation in analysing food, tourism, political and medical discourse. Her work has been cited in the creative industries, the hospitality sectors, as well as in international reports relating to the post-Covid recovery of the creative industries and global tourism. Her publications include monographs and refereed journal articles on themes ranging from menu translation to bi/trilingualism in secondary education. She is a co-editor of Perspectives: Studies in Translation Theory and Practice and a reviewer and panellist for AHRC, UKRI and several journals and publishers such as Routledge and Benjamin’s. She is also the general editor of Routledge Studies in Global Food Translation.