Responsible AI practice and AI education are central to AI implementation : a rapid review for all medical imaging professionals in Europe

Walsh, Gemma (author), Stogiannos, Nikolaos (author), Van de Venter, Riaan (author), Rainey, Claire (author), Tam, Winnie (author), McFadden, Sonyia (author), McNulty, Jonathan P. (author), Mekiš, Nejc (author), Lewis, Sarah (author), O'Regan, Tracy (author)

Objectives: It is essential to study the dosimetric performance and reliability of personal dosimeters. This study examines and compares the responses of two commercial thermoluminescence dosimeters (TLDs), the TLD-100 and the MTS-N. Methods: We compared the two TLDs to various parameters such as energy dependence, linearity, homogeneity, reproducibility, light sensitivity (zero point), angular dependence, and temperature effects using the IEC 61066 standard. Results: The results acquired showed that both TLD materials show linear behavior as indicated by the quality of the fit. In addition, the angular dependence results for both detectors show that all dose responses are within the range of acceptable values. However, the TLD-100 outperformed the MTS-N in terms of light sensitivity reproducibility for all detectors together, while the MTS-N outperforms the TLD-100 for each detector independently and that showed TLD-100 has more stability than MTS-N. The MTS-N shows better batch homogeneity (10.84%) than TLD-100 (13.65%). The effect of temperature in signal loss was clearer at higher temperature 65°C and it was however below ±30%. Conclusions: The overall results for dosimetric properties determined in terms of dose equivalents for all combinations of detectors are satisfactory. The MTS-N cards have better results in the energy dependence, angular dependency, batch homogeneity and less signal fading, whereas the TLD-100 cards are less sensitive to light and more reproducible. Advances in knowledge: Although previous studies showed several types of comparisons between TLDs, they have used limited parameters and different data analysis. This study has dealt with more comprehensive characterization methods and examinations combining TLD-100 and MTS-N cards.

• Language: English
• Keywords: artificial intelligence, healthcare, radiology, radiography, big data, ethical practices
• Work type: Article
• Typology: 1.02 - Review Article
• Organization: ZF - Faculty of Health Sciences
• Publication status: Published
• Publication version: Version of Record
• Publication date: 30.06.2023
• Year: 2023
• Number of pages: 9 str.
• Numbering: Vol. 5, iss. 1, art. 20230033
• PID: 20.500.12556/RUL-150069
• UDC: 616-07:004.8
• ISSN on article: 2513-9878
• DOI: 10.1259/bjro.20230033
• COBISS.SI-ID: 157921795
• Publication date in RUL: 13.09.2023

Record is a part of a journal

• Title: BJR|open
• Publisher: British Institute of Radiology
• ISSN: 2513-9878
• COBISS.SI-ID: 18870019

Licences

• License: CC BY 4.0, Creative Commons Attribution 4.0 International
• Link: http://creativecommons.org/licenses/by/4.0/
• Description: This is the standard Creative Commons license that gives others maximum freedom to do what they want with the work as long as they credit the author.
• Licensing start date: 30.06.2023

Secondary language

• Language: Undetermined
• Keywords: umetna inteligenca, zdravstvo, radiologija, radiografija, veliki podatki, etične prakse