Evaluation of chest tomosynthesis for the detection of pulmonary nodules: effect of clinical experience and comparison with chest radiography

Sara Zachrisson; Jenny Vikgren; Angelica Svalkvist; Åse A. Johnsson; Marianne Boijsen; Agneta Flinck; Lars Gunnar Månsson; Susanne Kheddache; Magnus Båth

doi:10.1117/12.811099

Translator Disclaimer

12 March 2009 Evaluation of chest tomosynthesis for the detection of pulmonary nodules: effect of clinical experience and comparison with chest radiography

Sara Zachrisson, Jenny Vikgren, Angelica Svalkvist, Åse A. Johnsson, Marianne Boijsen, Agneta Flinck, Lars Gunnar Månsson, Susanne Kheddache, Magnus Båth

Author Affiliations +

Proceedings Volume 7263, Medical Imaging 2009: Image Perception, Observer Performance, and Technology Assessment; 72630Z (2009) https://doi.org/10.1117/12.811099
Event: SPIE Medical Imaging, 2009, Lake Buena Vista (Orlando Area), Florida, United States

Abstract

Chest tomosynthesis refers to the technique of collecting low-dose projections of the chest at different angles and using these projections to reconstruct section images of the chest. In this study, a comparison of chest tomosynthesis and chest radiography in the detection of pulmonary nodules was performed and the effect of clinical experience of chest tomosynthesis was evaluated. Three senior thoracic radiologists, with more than ten years of experience of chest radiology and 6 months of clinical experience of chest tomosynthesis, acted as observers in a jackknife free-response receiver operating characteristics (JAFROC-1) study, performed on 42 patients with and 47 patients without pulmonary nodules examined with both chest tomosynthesis and chest radiography. MDCT was used as reference and the total number of nodules found using MDCT was 131. To investigate the effect of additional clinical experience of chest tomosynthesis, a second reading session of the tomosynthesis images was performed one year after the initial one. The JAFROC-1 figure of merit (FOM) was used as the principal measure of detectability. In comparison with chest radiography, chest tomosynthesis performed significantly better with regard to detectability. The observer-averaged JAFROC-1 FOM was 0.61 for tomosynthesis and 0.40 for radiography, giving a statistically significant difference between the techniques of 0.21 (p<0.0001). The observer-averaged JAFROC-1 FOM of the second reading of the tomosynthesis cases was not significantly higher than that of the first reading, indicating no improvement in detectability due to additional clinical experience of tomosynthesis.

Citation Download Citation

Sara Zachrisson, Jenny Vikgren, Angelica Svalkvist, Åse A. Johnsson, Marianne Boijsen, Agneta Flinck, Lars Gunnar Månsson, Susanne Kheddache, and Magnus Båth "Evaluation of chest tomosynthesis for the detection of pulmonary nodules: effect of clinical experience and comparison with chest radiography", Proc. SPIE 7263, Medical Imaging 2009: Image Perception, Observer Performance, and Technology Assessment, 72630Z (12 March 2009); https://doi.org/10.1117/12.811099

ACCESS THE FULL ARTICLE