Technical image quality assessment as a tool for clinical audit in interventional radiology: first results of a...

This study reports on a pilot clinical audit in interventional radiology (IR) in Italy, aimed at establishing a standardized framework for assessing technical image quality (TIQ) and patient dosimetry across X-ray systems. The initiative, supported by the Ministry of Health, aligns with European Directive 2013/59/Euratom and Continuous Quality Improvement principles.

Six medical physics centers applied a common protocol based on phantom setups. TIQ was assessed with two metrics: threshold contrast detail detectability (TCDD) using a statistical method (CTSM), and limiting spatial resolution (LSR) evaluated with the TOR 18FG phantom. Dosimetry included entrance surface Kerma per frame (ESK_frame) and Kerma rate (ESK_rate), measured with PMMA phantoms under cardiac, neurological, and abdominal protocols. A total of 26 fixed angiographic and 8 mobile C-arm systems were tested.

From 136 dosimetric sessions, 133 valid TCDD and 129 LSR datasets were obtained. Results showed wide variability across systems and protocols. Higher ESK_frame values generally improved TIQ metrics, though exceptions were observed, such as in DSA cine mode. LSR evaluations showed high inter-observer agreement (weighted Fleiss Kappa = 0.82). Systems with advanced noise-reduction algorithms demonstrated effective dose reduction with preserved image quality.

The audit confirms the feasibility of a standardized, reproducible approach to benchmarking IR equipment using phantom-based TIQ and dosimetry metrics. Variability among identical models highlights the influence of protocol customization and processing algorithms. Routine quality assessments are needed to ensure safety and optimize performance, with future efforts expanding system coverage and linking technical with clinical image quality criteria.