Local Time : 13:47 CET

C-3036 - An automated method to asses clinical image quality using FROC analysis for nodule detection in simulated lung phantom CT images

W. Veldkamp1, J. van der Vaart1, G. J. Streekstra2, I. Hernandez-Giron1; 1 Leiden/NL 2 Amsterdam/NL Type: Scientific Exhibit
Area of Interests: Thorax, Lung, Radiation physics
Imaging Techniques: Image manipulation / Reconstruction, CT-Quantitative, CT
Procedures: Computer Applications-General, Computer Applications-Detection, diagnosis, Physics
Special Focus: Quality assurance
Add poster to my schedule In your schedule (remove)

Direct link to access the poster (click here)

Aims and objectives: To develop a method for clinical image quality assessment in CT using FROC analysis.   As proof of concept, the detectability of lesions in simulated CT images of an anthropomorphic lung phantom was investigated.

Methods and materials: An algorithm to simulate lung vessels (diameters between 2.5mm and 8.5mm) in a phantom was generated (MATLAB) with lesion like spherical objects to stand for nodules (3mm diameter, -500HU) inserted in different locations. Ten lesions were simulated. Air was simulated inside and outside the lung (-90[...]

Results: FROC curves showed higher TPF for kernel_2 (lung standard) than for kernel_1 (soft body). FROC curves varied depending on the level of noise for both kernels, as expected, showing higher TPFs for lower noise.   For instance, in Fig. 4, it can be seen that for the highest noise level,[...]

Conclusion: An automated method to perform FROC analysis for nodule detection in lung phantom images was developed and tested in simulated CT data. This method can be applied to real scans of lung phantoms containing lesions.   Next step is a validation with human observer performance to investigate [...]

Personal information: W.J.H. Veldkamp, PhD*: Leiden University Medical Center (LUMC). Radiology Department. Albinusdreef 2, 2333ZA, Leiden, The Netherlands * email: W.J.H.Veldkamp@lumc.nl    J. van der Vaart: Leiden University Medical Center (LUMC). Radiology Department. Albinusdreef 2, 2333ZA, Leid[...]

References: 1. Chakraborty DP. A brief history of free-response receiver operating characteristic paradigm data analysis. Academic Radiology 20(7), 2013.   2. Verdun FR, Racine D, Ott JG, Tapiovaara MJ, Toroi P, et al. Image quality in CT: From physical measurements to model observers. Physica Medica[...]

This website uses cookies. Learn more