МЕТОДИКА ЭКСПЕРТНОЙ ОЦЕНКИ КАЧЕСТВА ПЭТ-ИЗОБРАЖЕНИЙ ПАЦИЕНТОВ

Abstract

Purpose: To develop the method of experts’ quality evaluation of the PET images as an additional quality control method for accurate, comparable, and reproducible PET diagnostics results, and to conduct image quality evaluation in different PET departments used this method. Material and methods: 60 PET images (without CT) of the patients who underwent whole body PET/CT with 18F-FDG were collected from 12 PET/CT scanners in 9 PET departments. Experts’ quality evaluation was conducted with questioning of the experts. Each expert evaluated the image quality by five-point scale and filled out the special form which include three image quality criteria: image clarity, artefacts, and general image quality. There were 28 experts from 8 different PET departments who have work experience in radiology from 1 to 32 years. The results of experts’ quality evaluation of the PET images were examined for correlations with parameters of acquisition and reconstruction protocols, examination methods. The results were also examined for dependance of subjective factors such as work experience and work conditions of experts. The minimum required number of experts were defined. The results were analyzed used statistical methods. Results: The PET images obtained by 8 PET/CT scanners had mean quality value more than 4 points (good quality). PET/CT scanners, which had the lowest quality value, have the obsolete or unusual settings and reconstruction parameters. The correlations between experts’ quality evaluation of the PET images and acquisition parameters (acquisition time per bed, multiplication of injected activity and acquisition time per bed), and examination methods (injected activity and uptake time) were established. The results of experts’ quality evaluation of the PET images were dependent on work experience and work conditions of experts. Conclusion: The method of experts’ quality evaluation of the PET images of the patients based on the questioning of the experts working in PET was developed and demonstrated in the current study. The results showed this method has the potential to compare the PET images obtained by different acquisition and reconstruction protocols, and it can be applied during the optimization of examination method and for the determination of obsolete and unusual settings of PET/CT. Experts’ evaluation of the PET images should include the opinion of at least six experts with different work experience in PET from several PET departments.