Comparison of detection of micro calcifications for Comparison of detection of micro calcifications for Comparison of detection of micro calcifications for Comparison of detection of micro calcifications for two clinical processing algorithms in digital two clinical processing algorithms in digital two clinical processing algorithms in digital two clinical processing algorithms in digital mammography mammography mammography mammography Abstract Abstract Abstract Abstract A potential advantage of digital radiology is the possibility to apply various clinical processing techniques. Clinical processing algorithms aim to better visualize the radiological image content. The experiment described in this paper is a quantitative method to compare the performance of 2 commercially available clinical processing algorithms for the detection of micro calcifications in digital mammograms. The first processing had been developed for general radiology. The second had been developed for general radiology and had been fine tuned for mammography. Digital mammograms from clinical practice were used. A specific number of small simulated micro calcifications with different sizes was inserted in the raw data. These composite images were then processed with the 2 different processing techniques. A reader study was conducted using hardcopy images. We report on the fraction of detected micro calcifications and False Positives (FP). The detected fractions were obvious higher with the processing that had been fine tuned for digital mammography. The number of FP was slightly higher with this processing algorithm. Findings in this study indicate that this evaluation method can be used to compare clinical processing protocols for the detection of micro calcifications. Additional tests should be applied to evaluate other aspects of processing such as detection of masses and evaluation of the global breast architecture. Introduction oduction oduction oduction One of the potential advantages of digital radiology is the possibility to apply image processing techniques. They can be applied at different stages in the imaging chain (figure 1)(1): 1. to correct detector artefacts 2. to enhance the image content by applying clinical processing protocols and 3. to prepare the image for display. The first type of processing is based on technical measurements such as flatfield images that serve for various calibration purposes. The corrections are then applied to all images regardless of their content. Image enhancement and software for better visualization aim to better visualize the radiological image content. In this study we have focused on the imaging processing that is applied on detector corrected clinical images. Figure 1: Overview of image data processing …
[1]
Perry Sprawls,et al.
The Expanding role of medical physics in diagnostic imaging
,
1997
.
[2]
Ann-Katherine Carton,et al.
Quantification of Al-equivalent thickness of just visible microcalcifications in full field digital mammograms.
,
2004,
Medical physics.
[3]
Arthur E. Burgess,et al.
Mammographic structure: data preparation and spatial statistics analysis
,
1999,
Medical Imaging.
[4]
K. Muller,et al.
Improving the detection of simulated masses in mammograms through two different image-processing techniques.
,
2001,
Academic radiology.
[5]
Andrew D. A. Maidment,et al.
Diagnostic accuracy of digital mammography in patients with dense breasts who underwent problem-solving mammography: effects of image processing and lesion type.
,
2003,
Radiology.
[6]
Development and validation of a simulation procedure to study the visibility of micro calcifications in digital mammograms.
,
2003,
Medical physics.
[7]
Robert M. Nishikawa,et al.
Radiologists’ Preferences for Digital Mammographic Display
,
2000
.
[8]
Harold L. Kundel,et al.
Handbook of Medical Imaging, Volume 1. Physics and Psychophysics
,
2000
.
[9]
James T. Dobbins.
Image Quality Metrics for Digital Systems
,
2000
.