Free Software for the Detection of Structures in Breast Images

Abstract:

The use of mammography as a method for the early detection of breast cancer reduces mortality from it. To aid in the diagnosis there are several free image processing software with the purpose of extracting characteristics such as microcalcification, nodules or edges of important structures. The objective of this study is to indicate which filters are best to aid in the processing of the images in three of these free software. The free software ImageJ, ImageTool and Mipav were used for the processing of 60 mammographic images from the Mini-MIAS database focusing the mean, median and Gaussian filters. The Haralick descriptors (second angular moment, contrast, entropy and correlation) and difference of values of gray levels (standard deviation, mean, minimum and maximum) were extracted from the resulting images. Then the results were compared taking into consideration the purpose of the processing. It was observed that it's best to use the mean and median filters from the Mipav software if the purpose of the processing is to leave the images with higher contrast levels. However, if the purpose is to obtain higher entropy levels, the ImageTool software should be used. It was also observed a similar processing time among the three software. The filter choice will depend on the type of noise to be removed from the image. For "Salt and Pepper" noise the mean filter should be used, while for the impulsive noise, the median one should be used. The results allowed the conclusion that the choice of the software to perform the processing of the mammographic images depends on the purpose of the processing application, if it's to increase the contrast in the image or if it's to extract other characteristics of diagnostic interest.

• Periodical: Applied Mechanics and Materials (Volume 590)
• Pages: 814-818
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.590.814
• Online since: June 2014
• Authors: Roger Campello de Magalhães, Silvia Cristina Martini Rodrigues, Márcia Aparecida Bissaco da Silva, Luiz Teruo Kawamoto*, Alessandro Pereira da Silva
• Keywords: Free Software, Image Processing, Mammographic Images

References

[1] INCA – Instituto Nacional do Câncer – Ministério da Saúde. Available online at: <http: / www. inca. gov. br >. Accessed: 13 Fev (2014).

[2] LANGARIZADEH, M., MAHMUD, R., RAMLI, A.R., NAPIS, S., BEIKZADEH, M.R., WAN ABDUL RAHMAN, W.E.Z., Effects of enhancement methods on diagnostic quality of digital mammogram images. Iranian Journal of Cancer Prevention, 3, 36–41, (2010).

[3] IMAGE TOOL - Available online at: <http: /ddsdx. uthscsa. edu/dig/itdesc. html>Accessed: 10 July (2013).

[4] IMAGEJ - Available online at: <http: /rsb. info. nih. gov/ij>Accessed: 13 August (2013).

[5] MIPAV – Available online at: <http: /mipav. cit. nih. gov/>Acessed: 23September (2013).

[6] MIAS DATABASE, Available online at: <http: /peipa. essex. ac. uk/ipa/pix/mias/>Acessed: 10 May (2013).

[7] SUCKLING, JK. ET AL The Mammographic Image Analysis Society DigitalMammogram Database, ExerptaMedica. International Congress Series 1069 pp.375-378, (1994).

[8] GONZALEZ, R.C. AND WOODS, R.E., 2002, Digital Image Processing, 2nd edn (Prentice Hall, New Jersey).