Diagnóstico de Câncer de Mama Através de Vetores de Descritores Localmente Agregados

• Ricardo Marques UFMA
• Geovane Ramos Neto UFMA
• Geraldo Braz Júnior UFMA
• João Almeida UFMA

DOI: https://doi.org/10.5753/sbcas.2016.9907

Abstract

An approach to early detect such breast anomalies is the mammography image. However, complex image patterns and the different organization of the breast tissues requires skill and experience by a trained physician to avoid faults in the mammograms interpretation. The main goal of this work is reduces the number of faults associateds to exam. For this, we propose a feature extraction of texture using Vector of Locally Aggregated Descriptors. Initial tests have achieved promising results , the better values obtained with 600 samples of DDSM base are: 90.18 (accuracy), 91.83 (sensitivity) and 94.02 (specificity).

References

Bay, H., Ess, A., Tuytelaars, T., and Van Gool, L. (2008). Speeded-up robust features (surf). Computer vision and image understanding, 110(3):346–359.

Gonzalez, R. andWoods, R. (2010). Processamento Digital de Imagens. Pearson Prentice Hall, S˜ao Paulo, 3 edition.

Heath, M., Bowyer, K., Kopans, D., and Moore, R. (2000). The digital database for screening mammography. Citeseer.

INCA (2016). Estimativas 2016: Incidência de Câncer no Brasil. http://http://www2.inca.gov.br/wps/wcm/connect/tiposdecancer/site/home/mama/.

Jégou, H., Douze, M., Schmid, C., and Pérez, P. (2010). Aggregating local descriptors into a compact image representation. In Computer Vision and Pattern Recognition (CVPR), 2010 IEEE Conference on, pages 3304–3311. IEEE.

Thornton, C., Hutter, F., Hoos, H. H., and Leyton-Brown, K. (2013). Auto-weka: Combined selection and hyperparameter optimization of classification algorithms. In Proceedings of the 19th ACM SIGKDD international conference on Knowledge discovery and data mining, pages 847–855. ACM.