Automated analysis of the American College of Radiology mammographic accreditation phantom images

Abstract

A significant metric in federal mammography quality standards is the phantom image quality assessment. The present work seeks to demonstrate that automated image analyses for American College of Radiology (ACR) mammographic accreditation phantom (MAP) images may be performed by a computer with objectivity, once a human acceptance level has been established. Twelve MAP images were generated with different x-ray techniques and digitized. Nineteen medical physicists in diagnostic roles (five of which were specially trained in mammography) viewed the original film images under similar conditions and provided individual scores for each test object (fibrils, microcalcifications, and nodules). Fourier domain template matching, used for low-level processing, combined with derivative filters, for intermediate-level processing, provided translation and rotation-independent localization of the test objects in the MAP images. The visibility classification decision was modeled by a Bayesian classifer using threshold contrast. The 50% visibility contrast threshold established by the trained observers' responses were: fibrils 1.010, microcalcifications 1.156, and nodules 1.016. Using these values as an estimate of human observer performance and given the automated localization of test objects, six images were graded with the computer algorithm. In all but one instance, the algorithm scored the images the same as the diagnostic physicists. In the case where it did not, the margin of disagreement was 10% due to the fact that the human scoring did not allow for half-visible fibrils (agreement occurred for the other test objects). The implication from this is that an operator-independent, machine-based scoring of MAP images is feasible and could be used as a tool to help eliminate the effect of observer variability within the current system, given proper, consistent digitization is performed.

Computerized quantitative evaluation of mammographic accreditation phantom images

“…[29][30][31] Moreover, one of the issues of this study is the comparison with other methods. [15][16][17][18][19][20] A combination of our method and other methods would improve the computerized scoring performance for ACR mammographic phantom images. However, the comparison and the combination should be executed carefully and the works need considerable efforts.…”

“…Similarly, Dougherty 20 did not include fiber objects in his study. Brooks et al 19 used frequency domain template matching to automatically localize each test object and determined a contrast of the shape to estimate object visibility. Their scheme demonstrated a successful ACR phantom evaluation process to all test object groups and all sizes, although only six test images were used for the analysis.…”

Mammographic Quality Control Using Digital Image Processing Techniques

“…Little work has been done to automate quality control in mammographic facilities. Fast Fourier transform is used (Brooks et al, 1997) to establish some visibility criteria for the phantom test object. (Chakraborty et al, 1997) compares phantom images with a pattern image to obtain relations between some of the image parameters and the physical conditions in which the images have been obtained.…”

Indicadores de calidad en imágenes digitales en programas de control de calidad en mamografía

“…E) Rango dinámico y linealidad en la detección 44 3 Definida, de una forma más precisa, en el capítulo 4…”

“…K.W. Brooks et al [3] desarrollan un programa de calidad de la imagen basado en el establecimiento de unos criterios de visibilidad para los objetos de test del fantoma, dichos objetos se localizan utilizando una técnica basada en la transformada rápida de Fourier. D.P.…”