An evaluation of a checklist in musculoskeletal radiographic image interpretation when using Artificial Intelligence (AI) - Oct 2022

The use of Artificial Intelligence (AI) has expanded throughout the radiography profession and is increasingly used in image interpretation tasks. The concept brings with it challenges for its optimal use in the reporting environments. Human reliance on the technology and bias can cause errors to be made. A checklist, used in conjunction with the AI, to mitigate against such biases, may optimise the use of the AI technology and promote good decision hygiene. Previous studies have demonstrated that trust issues exist amongst radiologists and radiographers in both over-reliance (automation bias) and reluctance to use AI systems for decision support (Pesapane et al. 2018). This proposed study aims to test the effect of a checklist for radiographic image assessment when using AI assistance. Participants will interpret images which have been previously reported to avoid any new revelations or changes to patient treatment. Radiographers will initially interpret the images with the assistance of AI and then another set of images with the use of AI and the checklist. The images will be different to avoid the effect of memory, but they will be similar in pathology number and distribution. This research is important to reduce variation in the impact of AI amongst reporting radiographers and optimise the use of AI in clinical decision support in image interpretation.