As per a survey by Statista, the use of artificial intelligence (AI) in the healthcare industry was fully functional by 2021. Hence, the usage of artificial intelligence in veterinary medicine is growing across a range of applications such as imaging, disease predictions, cancer treatments, and radiology.
This article highlights various applications and fields where AI is used in veterinary medicine.
8 examples of artificial intelligence in veterinary medicine
1. AI to aid in the detection of Addison’s Disease in dogs
Addison’s disease is difficult to diagnose due to the range of symptoms displayed by the disease. It’s also known as the Great Imitator. General signs of Addison’s disease are poor appetite, gastroenteritis, inability to respond to stress, and subsequent loss of body conditions. This disease can go undetected for years together.
In order to detect this disease, veterinarians at the University of California, Davis School of Veterinary Medicine, have created an algorithm using AI to diagnose Addison’s disease in canines. Krystle Reagan and Chen Gilor of UC Davis teamed up with a computer and electrical engineer to design the program. The accuracy rate of this artificial intelligence algorithm is more than 99 percent.
When an ill dog is brought to a clinic, the basic tests requested are blood tests. Lack of hormones linked with Addison’s disease leads to irregularities in the tests, which may be confused with several other diseases. Therefore, the team used the blood tests to train AI for identifying complex patterns from over 1,000 dogs that were treated at Davis.
The patterns are learned by the AI program to determine the disease. Then vets are alerted if the disease is detected, thus ultimately pushing them for further tests.
2. AI that helps vets to code their notes
Clinical records are essential resources for enhancing pet care and helping veterinarians with further research studies. However, vets lack the time to annotate clinical records with standard diagnostic codes, and most of the visits are recorded as text notes. The absence of a standard coding structure for using data creates an obstacle for improving care.
Therefore, to decrease the coding burden on vets, James Zou, PhD, assistant professor of biomedical data science and his team designed an AI solution called DeepTag. This algorithm reads the typed notes and estimates diseases.
The program breaks the notes into codes that point out symptoms, ailments, or diseases. DeepTag is trained on a database of 112,558 notes which are manually annotated.
3. Predicting chronic kidney disease (CKD) in Cats
Chronic kidney disease (CKD) symptoms in cats are bad breath, poor hair quality, weight loss, and lack of appetite. Vets are able to diagnose this disease with the IRIS (The International Renal Interest Society) system. However, artificial intelligence is aiding vets to detect CKD faster.
The American College of Veterinary Internal Medicine displayed an algorithm that leveraged data from over 150,000 cats above 20 years to anticipate CKD even before it occurred. This algorithm carries out analysis of huge health data to predict and not diagnose diseases in cats.
The algorithm helps vets to determine whether the cats will develop CKD in the next two years or not, with an accuracy level of 95 percent. While building the program, 35 data points were detected as possible estimators for the CKD.
In October 2019, Antech Diagnostics, based in South San Francisco, created a diagnostic tool using AI for predicting CKD in cats for up to two years. The algorithm analyzes the medical record of pets and anticipates the disease. This is a new way for vets to provide proactive care to cats.
4. AI in veterinary pathology
Pathologists need time and focus to scrutinize slides under a microscope and work on cell count along with measurements and calculations themselves. This work is time-consuming, can lead to errors, and is subjective. Therefore, in pathology AI programs can interpret accurate results and aid in decision making.
AI programs are taught to quantify, identify, measure, and collect information from microscopic images. Ready-to-use artificial intelligence tools and WSI (Whole Slide Imaging) for image analysis allows pathologists to examine slides effectively.
3 cases where AI is used in veterinary pathology:
1. Cell types and brain areas
Artificial intelligence image analysis is utilized in cellular neuroscience to check tissues on microscopic slides. This program is used to count certain cell types and measure brain areas. AI is programmed to calculate 1000s of cells in a minute.
2. Screening bone marrow cellularity changes
Analytical studies of bone marrow cellularity are represented by images. These pictures are unambiguous and interpretation of these is affected by the skillset of the researcher. AI programs offer quantification of bone marrow cellularity and enhance image analysis by scouting a larger area. This also provides precise cell count as compared to a pathologist.
3. Cutaneous mast cell tumor
Ki67-index and Ki67 stainings help to estimate the survival in the cutaneous mast cell tumor dogs. Counting stained cells is time-consuming, hence the usage of AI-driven image analysis gives results in just a few seconds.
5. SignalPET suite of AI products
SignalPET is a leading startup that provides AI-based veterinary medical tools.
3 AI-powered solutions provided by SignalPET:
1. SignalRAY
This product uses AI technology and deep machine learning to analyze animal radiographs in real-time. It gives reliable outcomes on around 50 plus radiograph results in 10 minutes. This AI program renders consistent and data-driven results to vets, thus improving outcomes and care.
2. SignalSMILE
In June 2021, SignalSMILE was launched with the objective of implementing AI-driven veterinary dentistry. This product is designed to detect five dental pathologies, such as Periapical Lucency, Furcation Bone Loss, Resorptive Lesion, tooth fracture, and Severe Attachment Loss (Alveolar Bone).
3. SignalPACS
SignalPACS is a Web PACS solution that’s created with cybersecurity in mind and with SignalPET’s artificial intelligence radiology software. This product has unlimited cloud storage capacity to meet today’s modern cybersecurity standards.
6. Vetology AI
Vetology AI provides teleradiology and artificial intelligence services, with clinically validated AI radiographs in just five minutes. Further, Vetology AI is based in California and was launched in 2017.
Their artificial intelligence tools are designed with the help of a technique called deep learning, which aids to detect abnormalities and diseases. AI training involves supplying hundreds of radiographs to the program, thus strengthening accuracy levels.
7. Cancer detection in the veterinary industry
According to the American Veterinary Medical Association, around half of the dogs above 10 years old will develop cancer. For cancer diagnosis and treatment, ImpriMed, a biotechnology company, checks cancer cells to estimate which chemotherapy will generate a response against tumor cells—helping owners and vets get a head start against cancer.
Live cancer cells are isolated from the tumor, then cultured in a laboratory. These cells are tested with FDA-approved chemotherapeutic agents that indicate cancer. ImpriMed gathers data from the live cancer cells of over 1,000 dogs every year. Effective treatments are anticipated with data analytics and cancer research, coupled with ImpriMed’s AI algorithm. This helps vets to decrease the time required to find the right chemotherapy and augments the database for improved care.
8. FidoCure
One Health, a life science startup company, launched FidoCure, a new program that combines artificial intelligence with genomics to render cancer treatments to canines. The integration of both helps to offer personalized cancer treatments. These insights are used by FidoCure for the treatment and research of human cancer cells too.
Canine genetic mutation is analyzed through gene sequencing for the cancer cells. FidoCure cooperates with the veterinary oncologist to detect mutations and find appropriate therapy for dogs.
3 applications of artificial intelligence in veterinary medicine
1. Data Analysis
Vets have access to surplus data acquired from patient records, research, devices, and other sources. These large data sets are difficult for humans to read and analyze, as humans are likely to miss out on essential aspects or make errors.
“AI algorithms are trained and programmed to read data, find correlations, destructure complex patterns, and check for abnormalities that are otherwise missed by vets,” says Erin Downes, VMD, Paoli Vetcare.
Expert.ai, an American intelligence specialist, has introduced a natural language platform that’s created to scrutinize a wide range of data from sources.
2. Imaging
Imaging is an essential medical procedure for diagnosis, however, there aren’t many people to interpret the images. Two AI startups, Vetology AI and SignalPET, that were included in the article provide services to interpret radiology results within a few minutes.
The AI software is designed to compare current and previous images, analyze them, and prioritize them too. Further, veterinary radiologists are required for reading complicated patterns in images, but artificial intelligence is designed to streamline the process.
3. Precision Medicine
Vets use precision for anticipating a treatment or a disease before its onset which is based on therapies and tests. AI precision medicine for disease detection provides useful insights to vets such as type of disease and further course of medications.
For instance, artificial intelligence for cancer detection in canines as discussed in the above points.
Final Thoughts
Artificial intelligence has set its mark in veterinary medicine and it won’t be long before it spreads to other aspects of veterinary practice management. Vets have applied technology for animal treatments—from anesthesia machines to digital imaging—to improve animal care and health outcomes, and they are wise to continue investing in technology and embracing the advancements that come from artificial intelligence.
Source : https://vetsuccess.com/blog/8-examples-artificial-intelligence-veterinary-medicine/page/5/
