Comparison of Lung Cancer Lesion Detection Capability on Standard Dose and Low Dose Computed Tomography Capabilities: An In-House Phantom Study

Abstract

The use of Low-Dose Computed Tomography (LDCT) protocols has garnered significant attention, particularly in detecting cancerous lesions in high-risk populations. However, the drawback of low-dose CT protocols results in image noise. Solutions introduced, such as the use of reconstruction techniques, tend to be time-inefficient, complex, and costly. This paper aims to explain the design and construction of an approach for evaluating the quality of lung cancer lesion imaging that is adequate and easily implementable. In this study, a custom-designed in-house phantom is required to simulate lung cancer lesions. The in-house phantom was constructed from organ or tissue-equivalent materials and equipped with various Hounsfield Unit values and lesion diameter sizes, which were determined based on data from 73 patients, consisting of both males and females, using contrast. Scans were performed on the phantom using standard-dose and low-dose protocol parameters. The results demonstrated that the low-dose protocol was able to detect small lesions at lower radiation levels. The contrast difference is quite good with a Signal Difference to Noise Ratio (SDNR) value ≥ 5. The image was optimum with a relatively high Figure of Merit (FOM). Additionally, Noise Power Spectrum (NPS) measurements provided accurate results within a specific range of spatial frequencies.