This poster is published under an
open license. Please read the
disclaimer for further details.
Keywords:
Quality assurance, Screening, Computer Applications-Detection, diagnosis, CAD, CT, Lung
Authors:
J. C. Marasinghe1, M. Ohkubo1, A. yoshida1, A. Kayugawa1, K. Murao2, T. Matsumoto3, S. Niizuma1, S. Sone 4, S. Wada1; 1Niigata/JP, 2Tokyo/JP, 3Chiba/JP, 4Nagano/JP
DOI:
10.1594/ecr2015/C-1125
Conclusion
Results indicate that QA of CAD will be required when changing the imaging system hardware/software components or image acquisition parameters.
Proposed QA protocol applies its own images of the CT scanner for QA process and more confidence results have obtained than applying different scanners’ images such as public databases’ images.
Furthermore nodule simulation is a solution for the burden of collecting larger number of nodules by an individual clinic.
Therefore,
additional cost and the time required to collect larger number of nodules can be reduced by applying reference virtual nodule set.
This study has some limitations.
Firstly,
only one CAD system has used for the study and the some other CAD systems with different algorithms also required to check the feasibility of the method.
Reference nodule set is simulated applying a spherical and uniform density object function.
However,
our idea is basic nodules are sufficient for QA process.
Proposed PSF-based virtual nodule method has demonstrated that the need of QA for lung cancer CT screening CAD systems by showing the CAD performance variation with different imaging system hardware/software components or image acquisition parameters.
Therefore,
this method can be potentially applied as a part of QA procedure for lung CAD.