5 Ways in which AI will Revolutionize Animal Health

Exploring Artificial Intelligence (AI) has oft been compared to the opening of Pandora’s Box. This famous Greek Mythological symbol of curiosity, for those who are unfamiliar, comes from the woman who opened a forbidden box and unleashed all the horrors of the earth upon humanity. “Pandora’s Box” now represents anything that is best left untouched, for fear of what might come out of it.

However, artificial intelligence should be seen as less of a box of horrors and more of a door to possibilities. In the past few months, the myriad uses of AI have rocked the internet. From internet influencers expounding on its merits to get follows and likes to content creators like myself using it to enhance their productions.

But artificial intelligence is not just a cute tool to get work done faster. Pandora let out death; artificial intelligence could help extend life. Pandora released disease; artificial intelligence could help diminish or even eradicate it. Pandora’s box contained poverty; artificial intelligence can release many from it.

In this article we will discuss the use cases and potential of artificial intelligence in the animal health sphere.

AI in Food & Agriculture

AI is being used for agricultural purposes. The world’s population is set to reach around 10 billion by 2050. With the amount of land remaining constant, an increased supply of food is dependent on changes in farming and animal rearing techniques and the fertility and yield of the land and cattle.

Identifying Animal Signatures

Contactless biometric AI (Artificial Intelligence) recognition tools, such as the one offered by F4T Lab, can create a ID by scanning the animal’s face. This eliminates the need for invasive and cumbersome hardware. It also removes the possibility of disputes after recovery from theft.

Data Gathering

Agricultural models can both gather large amounts of granular data, from genealogy, medical treatments, vaccines, and productivity. Evaluating farming methods, AI can offer smart solutions for better outcomes, minimized losses, and reduced workloads ^[[1]].

AI technology is being leveraged to regulate the usage of water, and systems that help track individual performance of every animal, from first parity until culling phase.

Herd Monitoring and Healthcare

AI-integrated UAVs can use optics and radiometric sensors to monitor and analyze special and temporal data on cattle location, movement, and interaction ^[[2]][[3]] that can perform tasks such as:

• Checking their current physical condition
• Determining which animals are high-performing
• Measuring stress levels and how they can affect production and fertility
• Getting better financial and insurance options because of more accurate and reliable animal tracking.
• Faster data sampling: Precision Hawk’s agricultural drones can “gather data on 500 to 1,000 acres in less than a day.”

In many cases, particular animal movements can suggest diseases, medical conditions, and weakness. If the values collected exceed the regular parameters, the Artificial Intelligence can be trained via deep learning to recognize this as ill health or injury. It can then alert animal caretakers to injuries, excessive weight, or the need for special attention.

Optimizing Breeding Strategies

Machine learning is having a huge impact on the optimization of animal genetic selection strategies. Various ML models exist that can aid in the prompt and efficient genetic selection of animals based on the forecasting of breeding values ^[[4]]­. Based on the geographical parameters, the system can provide farmers with recommendations for suitable and profitable breeds.

Mitigating Extreme Climatic Disruptions

The constantly shifting climate conditions have obfuscated the decision-making process for farmers and animal rearers. AI models can process a variety of factors and forecasted conditions to perform efficient water and environmental regulation. AI has been proven capable of sensing and reacting to the changes in heat by reducing ambient temperatures, aiding in heat loss e.g., shading and sheltering, and dietary alterations to reduce heat stress effects ^[[5]]. This function has proved accurate, with low statistical errors, ensuring high performance despite meteorological inconsistencies ^[[6]].

AI for Behavioral Sciences

Behavioral science is the study of influencing or maneuvering a creature’s behavioral patterns toward activities they might not inherently perform. Behavioral interventions are often used to modify negative behaviors.

Information Processing

AI can be useful in the compilation and synthesis of report findings in behavioral change intervention. While irregular and esoteric language is often used in these reports, machine learning coupled with advanced matching technology can interpret the information. This kind of AI would be immensely useful in predicting potential outcomes of behavioral interventions. AI can also perform personal profiling, which can be used to identify more volatile and aggressive herd members, forecast their misadventures, and prevent them from engaging in them.

Compliance

In a sea of treatments that must often be shoved down patients’ throats, nudgeomics is considered by some to be the best option. Nudge Theory uses the shifts in the environment, known as choice architecture, to direct or “nudge” the subjects towards certain actions. Combined with artificial intelligence, apps like DnaNudge can recommend genetically personalized nutritional products to raise patient compliance levels and reduce troubling behaviors. Furthermore, AI can determine based on past records and reports which behavioral technique (habituation, desensitization, counterconditioning, response substitution, overlearning, etc.) should be employed.

AI Diagnostic Capabilities

Recently, Deep Neural Networks (DNNs) have intensely propagated the pattern-recognition abilities of our AI technology. DNNs are data processing systems that mimic the brain’s learning style to assign labels to different sets of data. Simply put, if a machine is fed enough data on one particular condition, it can recognize it in any given scenario. This translates to artificial intelligence being highly capable of recognizing diseases from fresh scan images based on past learning.

Alzheimer’s

Cognitive decline and neuropathology are aspects of Alzheimer’s that dogs are able to develop. Typically, blood tests, X-rays, thyroid testing, and ultrasounds are used to confirm the disease. Recently, however, An AI-integrated histopathologic tool has been developed which opens a new paradigm for the study and diagnosis of Alzheimer’s brain disease ^[[7]]. By assessing the arrangements and features of the medial temporal lobe and the frontal cortex on slide images of brain autopsies, the algorithm was able to measure brain impairment to a substantial degree. The artificial intelligence-based biotech company, BERG, has developed applications to produce detailed disease maps that can help in the identification of disease biomarkers. Drugs can then be developed to target their sources.

Tuberculosis (TB)

Tuberculosis is an infectious disease that effects mammals such as goats, pigs, cats, and dogs, among others. Combined with nanotechnology, highly sensitive AI can enhance the diagnostic process of tuberculosis. By identifying the morphological changes caused by certain histological reactions, TB-AI can automatically detect acid-fast stained TB bacilli ^[[8]]. The nanotechnological component of the test can discern the minuscule molecules of the bacterial structure. When the immune cells envelop the bacteria, they shed what is essentially cell waste covered in Lymphangioleiomyomatosis, more succinctly known as LAM, and the protein LprG. The nanoparticles isolate these two molecules, confirming the presence of the TB bacteria.

Cardiological Conditions

The echocardiograph is a primary method of measuring the left ventricular ejection fraction (LVEF) and global longitudinal strain (GLS). These are the two most crucial indicators of cardiac systolic function. The AI algorithm, through data mining and interpretation, was able to analyze high-dimensional and complex data to identify hypertrophic cardiomyopathy with a sensitivity of 0.96 and specificity of 0.77 (Narula S. et al.). Another artificially intelligent program for associative memory classification was developed in a study by Sengupta P.P. et al. Here, the algorithm effectively distinguished constrictive pericarditis and restrictive cardiomyopathy (AUC 0.96).

An AI-enhanced echocardiogram was also able to detect heart failure more accurately than a blood test (NT-ProBNP) ^[[9]]. It would check for decreased heart function in patients with shortness of breath. A regular ECG can establish that there is a cardiovascular abnormality but this does not necessarily indicate heart failure. An ECG also tests for unusual BNP hormone levels in your blood but these can be symptomatic of a myriad of issues ranging from obesity to severe pulmonary hypertension. By training AI programs to recognize the ECG patterns of patients with LVSD, the researchers were able to receive highly accurate ECG analysis within about 10 seconds. Furthermore, deep learning algorithms can use coronary angiogram results to perform CAD tests that detect heart disease.

COVID-19

AI is currently being trained in University of Florida-supported research ^[][10] to detect fresh variants of COVID-19. This is intended to help tackle the onerous challenge of tracking the constantly mutating virus and to mitigate the most serious consequences. According to Marco Salemi, Ph.D., a professor of experimental pathology at the UF College of Medicine, “The coronavirus is a moving target and we have always been one step behind. Every time the epidemic seems to be coming under control, another variant emerges that is more virulent.” But with global sources of public data on the genetic sequences of coronavirus, preventative measures can be set in place to actively hinder its transmission.

Post-chemo Treatment

An AI-based system used in a study by the University of Michigan Health lab ^[[11]] can also determine the kind of treatment human bladder cancer patients will require post-chemotherapy. It is often very taxing to decipher the difference between cancer cells and scarred tissue. AI was able to significantly improve the evaluations of the trained practitioners, as well as provide a learning experience for the medical and veterinary students who participated in the study.

One interesting takeaway from the above study was that the AI tools could act as a second opinion, to aid a radiologist but not replace them. AI errors are distinct from human errors, so the two resources would need to be paired to cancel out each other’s shortcomings.

The use of AI in diagnosis procedures has been explored for the last two decades and at this stage, the uses are extensive. As more research and development occurs, the likelihood is that every day medical equipment and instruments will have built-in AI functions. Yet the question stands as to whether specialists will use these capabilities or discard them due to incompatibility with their systems or incomplete functions.

AI In Drug Development

There exists an abundance of significant clinical trial data in the databases of large pharmaceutical companies, yet much of it is in the form of unstructured, unsearchable data. NLP is an effective solution to this predicament. NLP (Natural Language Processing) is an Artificial Intelligence-based program that scans and converts images and unstructured texts into detailed and high-value, information-rich text. This enhances the precision of search results which can then be analyzed by the AI tech platform to glean insights and discover hidden implications.

BioXcel Therapeutics uses artificial intelligence technology to identify potential patients or new applications for existing drugs in the immuno-oncology and neuroscience fields. Deep Genomics is another biotech firm that has developed an AI platform to connect pharmaceutical researchers with the right test subjects and participants to maximize the likelihood of valid and reliable clinical endpoints. AI may also be involved in the trial design, recognizing risks and opportunities, forecasting trial duration, and analyzing the sequence of actions using associative or observational learning.

Hence, AI significantly aids in the designing and managing of clinical trials and data extraction. In doing so, the overall cost of preparing for, conducting, and evaluating the trial is neatly minimized and the drug release and distribution period is shortened.

AI Enhanced Communication Tools

Artificial intelligence can not only make calculations that exceed human capacity but also help people navigate through their devices with ease. This has been discovered with the use of Alexa, Siri, Cortana, and other AI personal assistants. Navigating medical devices and applications can also prove taxing. But having an AI guide and task-executer to help both veterinary students and clients adapt to new equipment and technological changes.

Implementing conversational AI into hospital services can enhance the client experience as well as customer relationship management. Buoy Health is an AI-powered chatbot that a client can explain health concern and symptoms to. The program can pinpoint key information and decipher whether the need is manageable or urgent. Based on its conclusion and their financial preference, it can then direct them to the appropriate medical center and treatment. AI chatbots have the potential to be used in complex medical consultations. When pet owners perceive chatbots as more accurate and easier to use, it increases their user satisfaction and behavioral intention to use the agents in veterinary consultations ^[[12]].

Botnation is a no-code solution to eliminating bias in veterinarians by simulating a pre-check-up interview session. The bot has been evaluated as having an “intuitive interface”, alongside flexible integration and conversation tracking. AI can augment health and self-management apps, providing predictive features and daily tracking of tests and readings, or facilitating engaging activities to assist with mental wellness. Sensely is a virtual assistant avatar that will visit its patient regularly, monitor their health, and ask their owners to examine their vital signs.

In agriculture, farmers and herders have used chatbots. These Artificial Intelligence powered algorithms assist them with unanswered questions ^[[13]], advising them on financially feasible products and equipment options, updating them on the latest erosion control strategies, and providing various other recommendations.

LIMITATIONS

While there are ostensibly no limitations for AI in the Life Sciences sectors, there are limitations to its adoption level. Many parties may reject or neglect the implementation or utilization of AI technology due to multiple reasons.

Animal owners might have inhibitions relegating the care of a valued family member to an artificial entity. Cattle rearers may be skeptical about providing AI software with access to their entire digital data archives. A 2019 study by Harvard Business Review showed that human patients were reluctant to use artificial intelligence software to receive healthcare. This is partially because people generally underestimate the accuracy and precision of AI-based diagnostic tools. They assume that AI would be ill-equipped to address their pet’s specific symptoms.

Meanwhile, as aforementioned, these have undergone much advancement to the point that the systems outperform specialists in their respective fields. Yet there is insufficient awareness of their scope coupled with skepticism emerging from misunderstandings surrounding the functionality of artificially intelligent algorithms. Patients may use them in conjunction with medical experts’ confirmations and supervision. Hence, it will take a while before the consumers of health services embrace this technology as a reliable and un-dismissible force in Life Sciences.

References

^[1] D.G. Panpatte, Artificial Intelligence in Agriculture: An Emerging Era of Research, Intuitional Science, CANADA (2018), pp. 1-8

[2] Alanezi, M.A.; Mohammad, A.; Sha’aban, Y.A.; Bouchekara, H.R.E.H.; Shahriar, M.S. Auto-Encoder Learning-Based UAV Communications for Livestock Management. Drones 2022, 6, 276. https://doi.org/10.3390/ drones6100276

[3] Rivas, A.; Chamoso, P.; González-Briones, A.; Corchado, J.M. Detection of cattle using drones and convolutional neural networks. Sensors 2018, 18, 2048.

[4] Hamadani, A., Ganai, N.A., Mudasir, S. et al. Comparison of artificial intelligence algorithms and their ranking for the prediction of genetic merit in sheep. Sci Rep 12, 18726 (2022). https://doi.org/10.1038/s41598-022-23499-w

[5] Fuentes, S.; Gonzalez Viejo, C.; Cullen, B.; Tongson, E.; Chauhan, S.S.; Dunshea, F.R. Artificial Intelligence Applied to a Robotic Dairy Farm to Model Milk Productivity and Quality based on Cow Data and Daily Environmental Parameters. Sensors 2020, 20, 2975. https://doi.org/10.3390/s20102975

^[6] Ahmed Elbeltagi, Nand Lal Kushwaha, Ankur Srivastava, Amira Talaat Zoof,

Chapter 5 – Artificial intelligent-based water and soil management, Deep Learning for Sustainable Agriculture, Pages 129-142,

^[7] https://healthitanalytics.com/news/researchers-leverage-ai-to-detect-causes-of-alzheimers-disease

[8] XIONG, Y., BA, X., HOU, A., ZHANG, K., CHEN, L., LI, T.. Automatic detection of mycobacterium tuberculosis using artificial intelligence. Journal of Thoracic Disease, North America, 10, mar. 2018. Available at: <https://jtd.amegroups.com/article/view/19696>

^[9] https://www.ahajournals.org/doi/pdf/10.1161/CIRCEP.120.008437

^[10] https://healthitanalytics.com/news/florida-researchers-to-use-ai-to-track-coronavirus-variants

^[11] “Computerized Decision Support for Bladder Cancer Treatment Response Assessment in CT Urography: Effect on Diagnostic Accuracy in Multi-Institution Multi-Specialty Study,” Tomography. DOI: 10.3390/tomography8020054

[12] Duen-Huang Huang, Hao-En Chueh, Chatbot usage intention analysis: Veterinary consultation, Journal of Innovation & Knowledge, Volume 6, Issue 3, 2021, Pages 135-144, ISSN 2444-569X, https://doi.org/10.1016/j.jik.2020.09.002.

^[13] Tanha Talaviya, Dhara Shah, Nivedita Patel, Hiteshri Yagnik, Manan Shah, Implementation of artificial intelligence in agriculture for optimisation of irrigation and application of pesticides and herbicides, Artificial Intelligence in Agriculture, Volume 4, 2020, Pages 58-73, ISSN 2589-7217,