Feasibility Study of Lesion Detection Using Deformable Part Models in Breast Ultrasound Images
Detection of lesions in ultrasound imaging typically requires human analysis due to their complexity. Hence, computerized lesion detection methods could be used to help radiologists in this process due to the fact that an early detection reduces the death rate caused by breast cancer. In this paper we propose a first experiment of a feasibility study for adapting a generic object detection technique, Deformable Part Models (DPM), to detect lesions in breast US images without any kind of human supervision. This technique has been evaluated in different topics obtaining prominent results. Hence, we propose a first assessment of this methodology applied to lesion detection in US images. We used a data-set composed by 50 images, all from different patients (18 malignant lesions, 32 benign lesions and 50 healthy tissue regions). In terms of quantitative results for lesion detection, our proposal obtains a sensitivity of 82% with 0.51 false-positive detections per image and an A z value of 0.96, which proves the feasibility of the proposal.
KeywordsBreast cancer lesion detection ultrasound deformable part models
Unable to display preview. Download preview PDF.
- 2.Dalal, N., Triggs, B.: Histograms of oriented gradients for human detection. In: Proceedings - 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, CVPR 2005, vol. I, pp. 886–893 (2005)Google Scholar
- 5.Everingham, M., Van Gool, L., Williams, C.K.I., Winn, J., Zisserman, A.: The PASCAL Visual Object Classes Challenge 2010 (VOC 2010) (2010), Results http://www.pascal-network.org/challenges/VOC/voc2010/workshop/index.html
- 8.Hao, Z., Wang, Q., Seong, Y.K., Lee, J.-H., Ren, H., Kim, J.-Y.: Combining CRF and multi-hypothesis detection for accurate lesion segmentation in breast sonograms. In: Ayache, N., Delingette, H., Golland, P., Mori, K. (eds.) MICCAI 2012, Part I. LNCS, vol. 7510, pp. 504–511. Springer, Heidelberg (2012)CrossRefGoogle Scholar
- 13.Stavros, A., Thickman, D., Rapp, C., Dennis, M., Parker, S., Sisney, G.: Solid breast nodules: Use of sonography to distinguish between benign and malignant lesions. Radiology 196(1), 123–134 (1995)Google Scholar