Chest X-Ray Imaging Algorithms for Optimising the Visualisation of Medical Lines and Tubes: A Study With Student...

Chest X-rays (CXRs) play a crucial role in determining the placement of medical lines and tubes. However, visualisation of these devices can be limited due to imaging and patient factors, increasing the risk of undetected misplacements, repeated radiation exposure and delayed patient management. These challenges may be more pronounced for student radiographers who are still developing their radiography skills. Although X-ray post-processing algorithms, such as grey-scale inversion (GSI) and Advanced Edge Enhancement (AEE), have the potential to optimise line and tube conspicuity, their effectiveness is not well established. Hence, this study aimed to evaluate the usefulness of GSI and AEE algorithms in enhancing visual image quality and reader confidence among student radiographers.

Twenty standard CXRs, demonstrating lines or tubes, were copied and processed using GSI and AEE algorithms. Images were paired as follows: Setting 1 (Standard-GSI), Setting 2 (Standard-AEE), Setting 3 (GSI-AEE). In each setting, participants used a relative visual grading characteristics (VGC) assessment to compare algorithms and indicated reader confidence using a 5-point Likert scale.

Twenty-five students completed the VGC assessment. The Friedman test revealed a significant increase in reader confidence (χ² (3) = 48.02, p < 0.001) in Settings 2 and 3. VGC analysis showed significantly higher visual grading scores with both GSI (AUC_VCG = 0.72, p < 0.001) and AEE (AUC_VCG = 0.91, p < 0.001) compared with the standard CXRs, with AEE outperforming GSI.

GSI and AEE appear to be valuable tools that merit integration into imaging protocols to promote best practices by optimising the use of resources, supporting diagnostic accuracy and potentially helping to minimise radiation exposure.