65. Study of respiratory signals of patients undergoing 4DCT for VMAT lung cancer treatments. (December 2018)

• Record Type: Journal Article
• Authors: Redapi, L.; Querci, M.; Marrazzo, L.; Calusi, S.; Talamonti, C.; Casati, M.; Pallotta, S.
• Is Part Of: Physica medica. Volume 56(2018)Supplement 2
• Page Start: 103
• Page End: 104
• Publication Date: 2018-12
• Subjects: Medical physics -- Periodicals; Biophysics -- Periodicals; Biophysics -- Periodicals; Imagerie médicale -- Périodiques; Radiothérapie -- Périodiques; Rayons X -- Sécurité -- Mesures -- Périodiques; Physique -- Périodiques; Médecine -- Périodiques; 610.153
• DOI: 10.1016/j.ejmp.2018.04.075
• Languages: English
• ISSNs: 1120-1797

Abstract

Abstract : Purpose: To account for respiratory motion in radiotherapy treatment of lung lesions, 4DCT is often employed. For a correct reconstruction of respiratory phases, periodic and regular breathing is the underlying hypothesis when phase-binning algorithms are used. Aim of this work was to study the respiratory signals acquired during 4DCT and to evaluate whether a correlation exists between deviations from regularity of breathing signal and quality of reconstructed images. Methods and materials: 4DCT of 40 patients acquired with a Philips Brilliance Big Bore CT were analyzed in this study. Respiratory signals were exported and an in-house written Matlab routine was used for extracting and processing the waveforms. The duration of each respiratory cycle (CD) was calculated as the distance between two consecutive maxima, while the fraction of inspiration (ID) was calculated as the ratio between the distance minimum–maximum and the duration of the whole cycle. Mean value (M), standard deviation ( σ ) and relative percentage error Er = σ M × 100 were extracted for each patient, the latter being a measurement of breathing regularity. Thereafter, 4DCT image quality was blindly scored from 1 to 4 (1: presence of visible artifacts in the reconstructed phases, 4: absence of visible artifacts). Patients were classified according to the score and mean relative percentage errors and standard deviations were calculated within each scoring class. Results: In Fig. 1 mean relative percentage errors versus assigned score are reported for both CD and ID. A correlation between Er and image quality is clearly visible. Threshold values for Er of about 18% for CD and about 15% for ID can be identified as limits between good and bad quality reconstructed images. Conclusion: Variations in respiratory cycle and inspiration phase duration clearly correlate with the quality of 4DCT reconstruction and can be used as predictors of image quality.