Veterinary medicine is a broad and growing discipline that includes topics such as companion animal health, population medicine and zoonotic diseases, and agriculture. In this article, we provide insight on how artificial intelligence works and how it is currently applied in veterinary medicine. We also discuss its potential in veterinary medicine. Given the rapid pace of research and commercial product developments in this area, the next several years will pose challenges to understanding, interpreting, and adopting this powerful and evolving technology. Artificial intelligence has the potential to enable veterinarians to perform tasks more efficiently while providing new insights for the management and treatment of disorders. It is our hope that this will translate to better quality of life for animals and those who care for them.
Artificial intelligence (AI) is a branch of computer science in which computer systems are designed to perform tasks that mimic human intelligence. Today, AI is reshaping day-to-day life and has numerous emerging medical applications poised to profoundly reshape the practice of veterinary medicine. In this Currents in One Health, we discuss the essential elements of AI for veterinary practitioners with the aim to help them make informed decisions in applying AI technologies into their practices. Veterinarians will play an integral role in ensuring the appropriate uses and good curation of data. The expertise of veterinary professionals will be vital to ensuring good data and, subsequently, AI that meets the needs of the profession. Readers interested in an in-depth description of AI and veterinary medicine are invited to explore a complementary manuscript of this Currents in One Health available in the May 2022 issue of the American Journal of Veterinary Research.
To identify potential risk factors for death following IV or intraosseous (IO) administration of contrast medium in birds undergoing CT scans.
120 birds that underwent 134 contrast-enhanced CT scans.
Medical records of birds of any species that underwent a CT scan which included administration of nonionic iodinated contrast medium from June 2013 to February 2020 were included. Information on birds and use of contrast medium was extracted from the medical records as well as information on deaths following IV or IO administration of contrast medium.
6 birds died shortly following administration of contrast medium. Necropsies were performed in 3 birds (2 cockatiels and 1 macaw), and all had lesions associated with the respiratory tract. When body weight was used as a binary variable to compare odds of death between small birds (≤ 150 g [0.33 lb]) and large birds (> 150 g), small birds had a 97-fold increased odds (OR, 97.5; 95% CI, 9.8 to 966.0) of dying following contrast medium administration. Following 131 CT scans with contrast medium administration (3 scans were excluded because of perivascular or subcutaneous leakage of contract medium), small birds had a mortality rate of 45.4% (5/11), compared with a mortality rate of 0.8% (1/120) for large (> 150 g) birds. Other variables (ie, sex, age, anesthesia or sedation, sedation protocol, and type of contrast medium) were not significantly associated with death after contrast medium administration.
CONCLUSIONS AND CLINICAL RELEVANCE
Although the administration of contrast medium cannot be conclusively confirmed as the cause of death in these birds, the high mortality rate for small birds coupled with the temporality of the event following contrast medium administration justifies the cautious use of contrast medium in small sick psittacine birds. (J Am Vet Med Assoc 2021;259:77–83)
To determine the feasibility of machine learning algorithms for the classification of appropriate collimation of the cranial and caudal borders in ventrodorsal and dorsoventral thoracic radiographs.
900 ventrodorsal and dorsoventral canine and feline thoracic radiographs were retrospectively acquired from the Picture Archiving and Communication system (PACs) system of the Ontario Veterinary College.
Radiographs acquired from April 2020 to May 2021 were labeled by 1 radiologist in Summer of 2022 as either appropriately or inappropriately collimated for the cranial and caudal borders. A machine learning model was trained to identify the appropriate inclusion of the entire lung field at both the cranial and caudal borders. Both individual models and a combined overall inclusion model were assessed based on the combined results of both the cranial and caudal border assessments.
The combined overall inclusion model showed a precision of 91.21% (95% CI [91, 91.4]), accuracy of 83.17% (95% CI [83, 83.4]), and F1 score of 87% (95% CI [86.8, 87.2]) for classification when compared with the radiologist’s quality assessment. The model took on average 6 ± 1 second to run.
Deep learning-based methods can classify small animal thoracic radiographs as appropriately or inappropriately collimated. These methods could be deployed in a clinical setting to improve the diagnostic quality of thoracic radiographs in small animal practice.
To describe the clinical presentation, treatment, and treatment outcomes for companion rats (Rattus norvegicus) diagnosed with lymphoma.
All rats that presented to the exotics service and underwent postmortem examination during the time period of 2008 through 2020 were evaluated.
The medical records of 35 rats were evaluated for an ante- or postmortem diagnosis of lymphoma. Cases with a diagnosis of lymphoma were further reviewed for signalment, presenting complaint, clinical signs observed on physical exam, diagnostic testing performed, and treatments administered. Postmortem gross and histologic findings were reviewed.
7 out of 35 rats were diagnosed with lymphoma, either ante-mortem or postmortem. The most common presenting complaint that was present in all rats with lymphoma was respiratory abnormalities. Five out of 7 rats had radiographs performed, all of which had abnormalities noted in the thoracic cavity including pulmonary nodules, cranial mediastinal widening, or alteration to the cardiac silhouette. Diagnosis via cytologic aspirates was performed in 2 cases and each was diagnostic for lymphoma; however, even with treatment, survival time following initiation of chemotherapy was short (less than or equal to 24 days). The definitive diagnosis in the remainder of the cases was via necropsy.
Results suggest that lymphoma is a common neoplastic disease in rats and a thorough diagnostic work-up is indicated in any rat that presents for general malaise or respiratory signs.
To report perioperative complications and client-perceived outcome following laparoscopic surgical treatment for sliding hiatal hernia (SHH) in dogs.
Client-owned dogs (n = 9).
Medical records were reviewed and perioperative data collected including preoperative diagnostic imaging, operative details, complications, and need for conversion to open celiotomy. A single-incision, multicannulated port was inserted in the subumbilical region followed by placement of an additional 2 or 3 instrument portals. Hiatal plication, esophagopexy, and left-sided gastropexy were performed laparoscopically. Follow-up information was collected with telephone interview with the owners and/or referring veterinarian. A standardized questionnaire was completed by dog owners postoperatively.
Intraoperative pneumothorax occurred in 5 of 9 (55.6%) dogs and resulted in conversion to open celiotomy in 2 (22.2%) dogs. In 4 dogs, pneumothorax was suspected to be the result of progressive leakage of capnoperitoneum through the suture bite holes of the esophageal hiatal plication sutures. Hiatal plication was performed using intracorporeal simple interrupted sutures (n = 4) or a simple continuous pattern with barbed suture (4). Esophagopexy was performed using barbed suture in all dogs. Gastropexy was performed using a total laparoscopic technique (n = 4) or laparoscopic-assisted technique (3). Using a standardized questionnaire, dog owners perceived a postoperative improvement in regurgitation after eating and regurgitation after excitement/increased activity.
Laparoscopic treatment of SHH resulted in owner-perceived improvement in clinical signs. Intraoperative pneumothorax occurred in a high proportion of dogs but did not result in long-term sequelae.
To report the perioperative characteristics and outcomes of dogs undergoing laparoscopic-assisted splenectomy (LAS).
136 client-owned dogs.
Multicentric retrospective study. Medical records of dogs undergoing LAS for treatment of naturally occurring splenic disease from January 1, 2014, to July 31, 2020, were reviewed. History, signalment, physical examination and preoperative diagnostic test results, procedural information, complications, duration of hospitalization, histopathologic diagnosis, and perioperative outcomes were recorded. Perioperative complications were defined using the Veterinary Cooperative Oncology Group – Common Terminology Criteria for Adverse Events (VCOG-CTCAE v2) guidelines.
LAS was performed for treatment of a splenic mass (124/136 [91%]), immune-mediated disease (7/136 [5%]), splenomegaly (4/136 [3%]), or immune-mediated disease in conjunction with a splenic mass (1/136 [1%]). Median splenic mass size was 1.3 cm3/kg body weight. Conversion to open laparotomy occurred in 5.9% (8/136) of dogs. Complications occurred in 78 dogs, with all being grade 2 or lower. Median surgical time was 47 minutes, and median postoperative hospital stay was 28 hours. All but 1 dog survived to discharge, the exception being postoperative death due to a suspected portal vein thrombus.
In the dogs of this report, LAS was associated with low rates of major complications, morbidity, and mortality when performed for a variety of splenic pathologies. Minimally invasive surgeons can consider the LAS technique to perform total splenectomy in dogs without hemoabdomen and with spleens with modest-sized splenic masses up to 55.2 cm3/kg, with minimal rates of complications, morbidity, and mortality.