Softwarové nástroje pro zpracování a analýzu testů kontroly kvality zobrazovacích přístrojů v nukleární medicíně

Abstract

This work was designed as a comprehensive set of software proposals for the semi-automated processing and quantitative evaluation of daily quality control tests of CT modules on hybrid imaging systems PET/CT (Siemens Biograph mCT 330) and SPECT/CT (GE Optima CT 640 NM). A total of 85 CT images were analyzed, from which—after preprocessing and removal of duplicates—31 images were processed for PET/CT and 37 for SPECT/CT. Using automated 3D segmentation of a water phantom, key image quality parameters were measured: uniformity (Mean Error ≤ 0.7 HU), noise (σ ≤ 4.2 HU), spatial resolution (FWHM ≈ 3.8 mm, R ≈ 0.26 mm⁻¹), and three-dimensional uniformity (U₃D ≥ 0.98). The measured values fully complied with the requirements of IEC 61223-3-4 and NEMA NU standards. The proposed solution is modular and ready for expansion with fully automated procedures or integration into clinical practice, which should significantly simplify routine quality control and improve the detection of potential malfunctions in imaging system operation.