Medical Image Analysis

Breast cancer is the most common type of cancer among women. Early stage detection of breast cancer is important as it can substantially reduce mortality rates. Mammography is the imaging modality of choice for breast cancer screening, demonstrating proven benefits for women older than 40 years. The diagnosis of a lesion based on mammography findings is challenging, resulting in a high number of follow-up examinations that most of the time prove to be unnecessary.

Computer-aided diagnosis (CADx) systems offer substantial assistance to the decision-making process of radiologists. The goal of such systems is to reduce the effort required for the diagnosis of a lesion, as well as reduce the high number of false positives leading to costly and discomforting biopsies.

The proposed approach is based on Content-based Image Retrieval. It aims to offer decision in the context of similar retrieved images that belong to cases stored in a reference database, providing radiologists with visual aid, justifying the produced results, leading to increased confidence in incorporating CAD-cued results in decision making. The contribution of the proposed methodology is the utilization of a set of support vector machines (SVM) during the retrieval stage, capable of exploiting the margin and pathology type likelihood in input samples along with the incorporation of a new texture descriptor, called mHOG, targeted to capturing margin and core specific mass properties.

  • L. Tsochatzidis, K. Zagoris, N. Arikidis, A. Karahaliou, L. Costaridou, I. Pratikakis “Computer-aided diagnosis of mammographic masses based on a supervised content-based image retrieval approach”, Pattern Recognition, 71, (106-117), 2017
  • L. Tsochatzidis, A. Karahaliou, K. Zagoris, S. Skiadopoulos, N. Arikidis, L. Costaridou, I. Pratikakis “A two-stage svm-based mammographic cbir for cadx”, 18th International Conference on Medical Image Computing and Computer Assisting Intervention (MICCAI) Workshop on Breast Image Analysis (BIA), 2015, pp. 41–48.
  • L. Tsochatzidis, K. Zagoris, M. Savelonas, N. Papamarkos, I. Pratikakis, N. Arikidis, L. Costaridou “Microcalcification oriented content-based mammogram retrieval for breast cancer diagnosis”, IEEE International Conference Imaging Systems and Techniques (IST), 2014, pp. 257 - 262