Fundamental study of a novel image quality index for optimising image quality and dose in digital radiography systems

The lack of comprehensive image quality index in digital radiography systems is an issue that needs to be addressed. This paper proposes a novel image quality index and reports a fundamental study employing a phantom. This study introduced SNR_IQ, which is a novel image quality index developed through a novel analytical algorithm and validated by phantom experiments.

To validate the performance of the proposed index, i.e., SNR_IQ, we conducted a comparative analysis using the conventional image quality index SNR_CONV and EI values. Using different subject thicknesses, grids, and doses, the CDRAD phantoms and abdominal phantoms were imaged. Inverse of image quality figure (IQF_inv) and the average psychological measurement values obtained by visual assessment were analyzed, and then compared with image quality indices.

The SNR_IQ evaluation demonstrated a significantly strong correlation with IQF_inv (R² = 0.949) evaluation and visual assessments (R² = 0.943). The SNR_IQ is a promising alternative, which correlates strongly with genuine image quality assessments, and provides immediate feedback alongside acquired images. Conversely, the SNR_CONV was inadequate as a comprehensive image quality index. The EI value served limited clinical purposes related to dose assessment, and it was inadequate as a comprehensive image quality index.

Our proposed SNR_IQ was able to assess image quality more accurately than conventional image quality indices.

By incorporating SNR_IQ, which is superior to conventional image quality indices, into existing systems, optimization of both dose and image quality can be improved. This rectifies the EI value's limitations, supports image quality management for various patient profiles, and enables informed decisions on dose adjustment or re-imaging. Extending exposure indicators to include SNR_IQ values can enhance examinations.