Optical density analysis of standard phantom for mammography quality control

Background/Objectives: The objective of this study was to analyze optic density of dummy lesions in breast phantom by mammography and understand whether the objectivity of visual inspections on dummy lesions upon appropriate and inappropriate decisions was correct quantitatively. Methods/Statistical analysis: Study subjects were 165 phantom images in 74 hospitals nationwide that passed the test of special medical equipment by Korean Institute for Accreditation of Medical Image from May 2016 to April 2017. Min (A), Max (B), Mean (C), and StdDev (D) were measured using color level information. Fibers, specks, and masses divided dummy lesions determined as appropriate or inappropriate in 165 images. For each divided dummy lesion, Min, Max, Mean concentration, and concentration difference summarized it. Findings: There was a significant difference in Max concentration (B) between dummy lesions of 9 (specks 3) and 10 (specks 4). In dummy lesion 12 (mass 1), there was no significant difference by each step, although the deviation between black and white was high since the scope of lesion was big. Upon analysis of optic density divided by appropriate and inappropriate decisions, there were significant differences in concentration difference (D) for fibers, Min (A) and Max (B) concentrations as well as concentration difference (D) for specks, and Max (B) and Mean (C) concentrations with concentration difference (D) for masses. Improvements/Applications: Visual inspections appeared to have difficulty in analyzing lesions due to the ambiguity of quantitative differences. Further developments of quantitative programs are needed to replace visual inspection for breast phantom lesions in mammography.