Data Mining for Patient Safety: Comparative Evaluation of HRCT Radiation Dose in COVID-19 patients at Eastern Libya

Abstract

This study compared radiation doses from fixed tube current and automatic exposure control (AEC) HRCT protocols in COVID-19 patients at Albyda Medical Center, Libya, and validated the findings using a controlled phantom experiment. A retrospective analysis was performed on 76 examinations acquired between January and June 2022. Scans were obtained a Fujifilm FCT SPEEDA scanner operating with fixed tube current and a Toshiba Aquilion scanner using AEC. Radiation dose metrics volume CT dose index (CTDIvol) and dose-length product (DLP) were extracted from the DICOM headers. Scan length was calculated as DLP/CTDIvol. As dose variables were non-normally distributed (Shapiro–Wilk test), non-parametric analyses were used. A complementary phantom study on the fixed-mAs scanner assessed four protocol variants for potential dose reduction. The AEC-equipped scanner produced significantly lower doses than the fixed-mAs system, with median CTDIvol values of 6.3 versus 12.1 mGy (p < 0.001) and median DLP values of 220.5 versus 418.7 mGy•cm (p < 0.001), representing a 47.3% DLP reduction. Scan lengths were comparable between systems (p = 0.423), indicating that dose savings were attributable to AEC rather than shorter scan lengths. In the phantom study, progressive protocol optimization on the fixed-mAs scanner achieved up to a 45.7% reduction in DLP. Dose substantially reduction for COVID-19 HRCT without effecting on image quality nor scan-length considerable reduction. In resource-limited settings such as Libya, using AEC-equipped systems and optimizing existing CT protocols can help maintain the ALARA principle while preserving diagnostic performance.