The unmet potential of artificial intelligence in veterinary medicine

Artificial intelligence in veterinary diagnostic imaging: Perspectives and limitations

The field of veterinary diagnostic imaging is undergoing significant transformation with the integration of artificial intelligence (AI) tools. This manuscript provides an overview of the current state and future prospects of AI in veterinary diagnostic imaging. The manuscript delves into various applications of AI across different imaging modalities, such as radiology, ultrasound, computed tomography, and magnetic resonance imaging. Examples of AI applications in each modality are provided, ranging from orthopaedics to internal medicine, cardiology, and more. Notable studies are discussed, demonstrating AI's potential for improved accuracy in detecting and classifying various abnormalities. The ethical considerations of using AI in veterinary diagnostics are also explored, highlighting the need for transparent AI development, accurate training data, awareness of the limitations of AI models, and the importance of maintaining human expertise in the decision-making process. The manuscript underscores the significance of AI as a decision support tool rather than a replacement for human judgement. In conclusion, this comprehensive manuscript offers an assessment of the current landscape and future potential of AI in veterinary diagnostic imaging. It provides insights into the benefits and challenges of integrating AI into clinical practice while emphasizing the critical role of ethics and human expertise in ensuring the wellbeing of veterinary patients.

Surveyed veterinary students in Australia find ChatGPT practical and relevant while expressing no concern about artificial intelligence replacing veterinarians

Background

Chat Generative Pre-trained Transformer (ChatGPT) is a freely available online artificial intelligence (AI) program capable of understanding and generating human-like language. This study assessed veterinary students' perceptions about ChatGPT in education and practice. It compared perceptions about ChatGPT between students who had completed a critical analysis task and those who had not.

Methods

This cross-sectional study surveyed 498 Doctor of Veterinary Medicine (DVM) students at The University of Sydney, Australia. Second-year DVM students researched a veterinary pathogen and then completed a critical analysis of ChatGPT (version 3.5) output for the same pathogen. A survey based on the Technology Acceptance Model was then delivered to all DVM students from all years of the programme, collecting data using Likert-style, categorical and free-text items.

Results

Over 75% of the 100 respondents reported having used ChatGPT. The students found ChatGPT's output relevant and practical for their use but perceived it as inaccurate. They perceived ChatGPT output to be more useful for veterinary students than for pet owners or veterinarians. Those who had completed the critical analysis assignment had a more positive view of ChatGPT's practicality for veterinary students but noted its authoritative tone even when delivering inaccurate information. Over 50% of the students agreed that information about tools such as ChatGPT should be included in the veterinary curriculum. Students agreed that veterinarians should embrace AI but disagreed that AI would eventually replace the need for veterinarians.

Conclusions

A critical appraisal of outputs from AI tools such as ChatGPT may help prepare future veterinarians for the effective use of these tools.

AI chatbots in pet health care: Opportunities and challenges for owners

The integration of artificial intelligence (AI) into health care has seen remarkable advancements, with applications extending to animal health. This article explores the potential benefits and challenges associated with employing AI chatbots as tools for pet health care. Focusing on ChatGPT, a prominent language model, the authors elucidate its capabilities and its potential impact on pet owners' decision-making processes. AI chatbots offer pet owners access to extensive information on animal health, research studies and diagnostic options, providing a cost-effective and convenient alternative to traditional veterinary consultations. The fate of a case involving a Border Collie named Sassy demonstrates the potential benefits of AI in veterinary medicine. In this instance, ChatGPT played a pivotal role in suggesting a diagnosis that led to successful treatment, showcasing the potential of AI chatbots as valuable tools in complex cases. However, concerns arise regarding pet owners relying solely on AI chatbots for medical advice, potentially resulting in misdiagnosis, inappropriate treatment and delayed professional intervention. We emphasize the need for a balanced approach, positioning AI chatbots as supplementary tools rather than substitutes for licensed veterinarians. To mitigate risks, the article proposes strategies such as educating pet owners on AI chatbots' limitations, implementing regulations to guide AI chatbot companies and fostering collaboration between AI chatbots and veterinarians. The intricate web of responsibilities in this dynamic landscape underscores the importance of government regulations, the educational role of AI chatbots and the symbiotic relationship between AI technology and veterinary expertise. In conclusion, while AI chatbots hold immense promise in transforming pet health care, cautious and informed usage is crucial. By promoting awareness, establishing regulations and fostering collaboration, the article advocates for a responsible integration of AI chatbots to ensure optimal care for pets.

Use of machine learning and Poincaré density grid in the diagnosis of sinus node dysfunction caused by sinoatrial conduction block in dogs

BACKGROUND

Sinus node dysfunction because of abnormal impulse generation or sinoatrial conduction block causes bradycardia that can be difficult to differentiate from high parasympathetic/low sympathetic modulation (HP/LSM).

HYPOTHESIS

Beat-to-beat relationships of sinus node dysfunction are quantifiably distinguishable by Poincaré plots, machine learning, and 3-dimensional density grid analysis. Moreover, computer modeling establishes sinoatrial conduction block as a mechanism.

ANIMALS

Three groups of dogs were studied with a diagnosis of: (1) balanced autonomic modulation (n = 26), (2) HP/LSM (n = 26), and (3) sinus node dysfunction (n = 21).

METHODS

Heart rate parameters and Poincaré plot data were determined [median (25%-75%)]. Recordings were randomly assigned to training or testing. Supervised machine learning of the training data was evaluated with the testing data. The computer model included impulse rate, exit block probability, and HP/LSM.

RESULTS

Confusion matrices illustrated the effectiveness in diagnosing by both machine learning and Poincaré density grid. Sinus pauses >2 s differentiated (P < .0001) HP/LSM (2340; 583-3947 s) from sinus node dysfunction (8503; 7078-10 050 s), but average heart rate did not. The shortest linear intervals were longer with sinus node dysfunction (315; 278-323 ms) vs HP/LSM (260; 251-292 ms; P = .008), but the longest linear intervals were shorter with sinus node dysfunction (620; 565-698 ms) vs HP/LSM (843; 799-888 ms; P < .0001).

CONCLUSIONS

Number and duration of pauses, not heart rate, differentiated sinus node dysfunction from HP/LSM. Machine learning and Poincaré density grid can accurately identify sinus node dysfunction. Computer modeling supports sinoatrial conduction block as a mechanism of sinus node dysfunction.

Comparison of Machine Learning Tree-Based Algorithms to Predict Future Paratuberculosis ELISA Results Using Repeat Milk Tests

Machine learning algorithms have been applied to various animal husbandry and veterinary-related problems; however, its use in Johne's disease diagnosis and control is still in its infancy. The following proof-of-concept study explores the application of tree-based (decision trees and random forest) algorithms to analyze repeat milk testing data from 1197 Canadian dairy cows and the algorithms' ability to predict future Johne's test results. The random forest models using milk component testing results alongside past Johne's results demonstrated a good predictive performance for a future Johne's ELISA result with a dichotomous outcome (positive vs. negative). The final random forest model yielded a kappa of 0.626, a roc AUC of 0.915, a sensitivity of 72%, and a specificity of 98%. The positive predictive and negative predictive values were 0.81 and 0.97, respectively. The decision tree models provided an interpretable alternative to the random forest algorithms with a slight decrease in model sensitivity. The results of this research suggest a promising avenue for future targeted Johne's testing schemes. Further research is needed to validate these techniques in real-world settings and explore their incorporation in prevention and control programs.

Computer vision model for the detection of canine pododermatitis and neoplasia of the paw

BACKGROUND

Artificial intelligence (AI) has been used successfully in human dermatology. AI utilises convolutional neural networks (CNN) to accomplish tasks such as image classification, object detection and segmentation, facilitating early diagnosis. Computer vision (CV), a field of AI, has shown great results in detecting signs of human skin diseases. Canine paw skin diseases are a common problem in general veterinary practice, and computer vision tools could facilitate the detection and monitoring of disease processes. Currently, no such tool is available in veterinary dermatology.

ANIMALS

Digital images of paws from healthy dogs and paws with pododermatitis or neoplasia were used.

OBJECTIVES

We tested the novel object detection model Pawgnosis, a Tiny YOLOv4 image analysis model deployed on a microcomputer with a camera for the rapid detection of canine pododermatitis and neoplasia.

MATERIALS AND METHODS

The prediction performance metrics used to evaluate the models included mean average precision (mAP), precision, recall, average precision (AP) for accuracy and frames per second (FPS) for speed.

RESULTS

A large dataset labelled by a single individual (Dataset A) used to train a Tiny YOLOv4 model provided the best results with a mean mAP of 0.95, precision of 0.86, recall of 0.93 and 20 FPS.

CONCLUSIONS AND CLINICAL RELEVANCE

This novel object detection model has the potential for application in the field of veterinary dermatology.

Evaluation of a Novel Veterinary Dental Radiography Artificial Intelligence Software Program

There is a growing trend of artificial intelligence (AI) applications in veterinary medicine, with the potential to assist veterinarians in clinical decisions. A commercially available, AI-based software program (AISP) for detecting common radiographic dental pathologies in dogs and cats was assessed for agreement with two human evaluators. Furcation bone loss, periapical lucency, resorptive lesion, retained tooth root, attachment (alveolar bone) loss and tooth fracture were assessed. The AISP does not attempt to diagnose or provide treatment recommendations, nor has it been trained to identify other types of radiographic pathology. Inter-rater reliability for detecting pathologies was measured by absolute percent agreement and Gwet's agreement coefficient. There was good to excellent inter-rater reliability among all raters, suggesting the AISP performs similarly at detecting the specified pathologies compared to human evaluators. Sensitivity and specificity for the AISP were assessed using human evaluators as the reference standard. The results revealed a trend of low sensitivity and high specificity, suggesting the AISP may produce a high rate of false negatives and may not be a good tool for initial screening. However, the low rate of false positives produced by the AISP suggests it may be beneficial as a "second set of eyes" because if it detects the specific pathology, there is a high likelihood that the pathology is present. With an understanding of the AISP, as an aid and not a substitute for veterinarians, the technology may increase dental radiography utilization and diagnostic potential.

Training and validation of a novel non-invasive imaging system for ruling out malignancy in canine subcutaneous and cutaneous masses using machine learning in 664 masses

Objective

To train and validate the use of a novel artificial intelligence-based thermal imaging system as a screening tool to rule out malignancy in cutaneous and subcutaneous masses in dogs.

Animals

Training study: 147 client-owned dogs with 233 masses. Validation Study: 299 client-owned dogs with 525 masses. Cytology was non-diagnostic in 94 masses, resulting in 431 masses from 248 dogs with diagnostic samples.

Procedures

The prospective studies were conducted between June 2020 and July 2022. During the scan, each mass and its adjacent healthy tissue was heated by a high-power Light-Emitting Diode. The tissue temperature was recorded by the device and consequently analyzed using a supervised machine learning algorithm to determine whether the mass required further investigation. The first study was performed to collect data to train the algorithm. The second study validated the algorithm, as the real-time device predictions were compared to the cytology and/or biopsy results.

Results

The results for the validation study were that the device correctly classified 45 out of 53 malignant masses and 253 out of 378 benign masses (sensitivity = 85% and specificity = 67%). The negative predictive value of the system (i.e., percent of benign masses identified as benign) was 97%.

Clinical relevance

The results demonstrate that this novel system could be used as a decision-support tool at the point of care, enabling clinicians to differentiate between benign lesions and those requiring further diagnostics.

Machine learning-based risk prediction model for canine myxomatous mitral valve disease using electronic health record data

Introduction

Myxomatous mitral valve disease (MMVD) is the most common cause of heart failure in dogs, and assessing the risk of heart failure in dogs with MMVD is often challenging. Machine learning applied to electronic health records (EHRs) is an effective tool for predicting prognosis in the medical field. This study aimed to develop machine learning-based heart failure risk prediction models for dogs with MMVD using a dataset of EHRs.

Methods

A total of 143 dogs with MMVD between May 2018 and May 2022. Complete medical records were reviewed for all patients. Demographic data, radiographic measurements, echocardiographic values, and laboratory results were obtained from the clinical database. Four machine-learning algorithms (random forest, K-nearest neighbors, naïve Bayes, support vector machine) were used to develop risk prediction models. Model performance was represented by plotting the receiver operating characteristic (ROC) curve and calculating the area under the curve (AUC). The best-performing model was chosen for the feature-ranking process.

Results

The random forest model showed superior performance to the other models (AUC = 0.88), while the performance of the K-nearest neighbors model showed the lowest performance (AUC = 0.69). The top three models showed excellent performance (AUC ≥ 0.8). According to the random forest algorithm's feature ranking, echocardiographic and radiographic variables had the highest predictive values for heart failure, followed by packed cell volume (PCV) and respiratory rates. Among the electrolyte variables, chloride had the highest predictive value for heart failure.

Discussion

These machine-learning models will enable clinicians to support decision-making in estimating the prognosis of patients with MMVD.

Using Machine Learning in Veterinary Medical Education: An Introduction for Veterinary Medicine Educators

Machine learning (ML) offers potential opportunities to enhance the learning, teaching, and assessments within veterinary medical education including but not limited to assisting with admissions processes as well as student progress evaluations. The purpose of this primer is to assist veterinary educators in appraising and potentially adopting these rapid upcoming advances in data science and technology. In the first section, we introduce ML concepts and highlight similarities/differences between ML and classical statistics. In the second section, we provide a step-by-step worked example using simulated veterinary student data to answer a hypothesis-driven question. Python syntax with explanations is provided within the text to create a random forest ML prediction model, a model composed of decision trees with each decision tree being composed of nodes and leaves. Within each step of the model creation, specific considerations such as how to manage incomplete student records are highlighted when applying ML algorithms within the veterinary education field. The results from the simulated data demonstrate how decisions by the veterinary educator during ML model creation may impact the most important features contributing to the model. These results highlight the need for the veterinary educator to be fully transparent during the creation of ML models and future research is needed to establish guidelines for handling data not missing at random in medical education, and preferred methods for model evaluation.

Veterinarian burnout demographics and organizational impacts: a narrative review

Burnout is a work-related syndrome of physical and emotional exhaustion secondary to prolonged, unresolvable occupational stress. Individuals of different demographic cohorts may have disparate experiences of workplace stressors and burnout impacts. Healthcare organizations are adversely affected by burnt out workers through decreased productivity, low morale, suboptimal teamwork, and potential impacts on the quality of patient care. In this second of two companion reviews, the demographics of veterinary burnout and the impacts of burnout on affected individuals and work environments are summarized, before discussing mitigation concepts and their extrapolation for targeted strategies within the veterinary workplace and profession.

Automated Wound Image Segmentation: Transfer Learning from Human to Pet via Active Semi-Supervised Learning

Wound management is a fundamental task in standard clinical practice. Automated solutions already exist for humans, but there is a lack of applications regarding wound management for pets. Precise and efficient wound assessment is helpful to improve diagnosis and to increase the effectiveness of treatment plans for chronic wounds. In this work, we introduced a novel pipeline for the segmentation of pet wound images. Starting from a model pre-trained on human-based wound images, we applied a combination of transfer learning (TL) and active semi-supervised learning (ASSL) to automatically label a large dataset. Additionally, we provided a guideline for future applications of TL+ASSL training strategy on image datasets. We compared the effectiveness of the proposed training strategy, monitoring the performance of an EfficientNet-b3 U-Net model against the lighter solution provided by a MobileNet-v2 U-Net model. We obtained 80% of correctly segmented images after five rounds of ASSL training. The EfficientNet-b3 U-Net model significantly outperformed the MobileNet-v2 one. We proved that the number of available samples is a key factor for the correct usage of ASSL training. The proposed approach is a viable solution to reduce the time required for the generation of a segmentation dataset.

Classification of racehorse limb radiographs using deep convolutional neural networks

Purpose

To assess the capability of deep convolutional neural networks to classify anatomical location and projection from a series of 48 standard views of racehorse limbs.

Materials and methods

Radiographs (N = 9504) of horse limbs from image sets made for veterinary inspections by 10 independent veterinary clinics were used to train, validate and test (116, 40 and 42 radiographs, respectively) six deep learning architectures available as part of the open source machine learning framework PyTorch. The deep learning architectures with the best top-1 accuracy had the batch size further investigated.

Results

Top-1 accuracy of six deep learning architectures ranged from 0.737 to 0.841. Top-1 accuracy of the best deep learning architecture (ResNet-34) ranged from 0.809 to 0.878, depending on batch size. ResNet-34 (batch size = 8) achieved the highest top-1 accuracy (0.878) and the majority (91.8%) of misclassification was due to laterality error. Class activation maps indicated that joint morphology, not side markers or other non-anatomical image regions, drove the model decision.

Conclusions

Deep convolutional neural networks can classify equine pre-import radiographs into the 48 standard views including moderate discrimination of laterality, independent of side marker presence.