University Library

Background: As Positron Emission Tomography use is increasing in veterinary medicine, personnel are potentially exposed to increasing levels of occupational radiation.

Objectives: To estimate the average exposure received by veterinary personnel during an F18-FDG PET procedure by reviewing the existing literature in both veterinary and human medicine. 

Design:  Primary and secondary literature, including grey literature (conference proceedings), was collected in English from 1978 from citation databases: CAB Abstracts (CABDirect), Scopus, CINAHL(Ebsco), and Biosis(Web of Science) on 12 January 2024 utilizing Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.  Data will be characterized from the study using tables and figures to describe protocol information, form of dosimetry, comparison/control populations, personnel role, animal species, average exposures, number of scans, duration of exposure, comparisons with other dosimetry or measures and reporting guidelines used in each study.

Large Language Models (LLMs) such as Open AI Chat GPT, Microsoft Bing, and Google Bard have received widespread coverage in the popular and higher education press. However, an important aspect of using these tools effectively has attracted minimal attention – how an individual phrases their query, or prompts the tool, impacts the content and quality of the response.  

Librarians have extensive experience helping students translate their research topics and questions into searchable queries across various online platforms. This poster includes an overview and assessment of existing AI tools and highlights what we have learned about crafting effective LLM prompts.

The emergence of new AI tools for information retrieval and synthesis is exciting; however, there are serious caveats. Instructors should encourage students to recognize the biases associated with using AI tools.