Research And Implementation Of Mammograms Information Platform

Breast cancer is one of the most common malignant diseases, and the early diagnosis and treatment can significantly increase the chance of survival for patients. Mammography is the first choice to diagnose breast cancer. However, features of early breast cancer in mammograms are unconspicuous, so even the seasoned doctor cannot easily discover all of the possible disease in time. With the rapid progress of physic and computer technology, computer aided detection (CAD) on mammograms has been a hot research field in early detection of breast cancer. A good CAD system, which can effectively avoid error and miss in diagnosis due to the eyestrain or negligence of human, could greatly facilitate the recognition of micro disease in medical images.In order to apply CAD into various fields, it is necessary to research not only the basic algorithm theory, but also the information platform development. The state of the art in CAD and related image segmentation algorithms are introduced and analyzed firstly, and then the design and implementation of the platform are discussed in detail. Finally, the pectoral muscle segmentation algorithm and the breast border detection algorithm based on the platform are presented. The contribution of this paper can be summarized as follows:1. By using design pattern and J2EE technology, we designed and accomplished a platform to meet the needs of clinical image management and medical college teaching;2. According to the features of the pectoral muscle in mammograms, a pectoral muscle segmentation algorithm is proposed, which can obtain the precise border of the pectoral muscle and segment the pectoral muscle region effectively;3. With the difficulties of the breast border detection being taken into account, an algorithm is presented to remove the artifact and extract the border based on morphological method.The medical information platform has been used in information management and teaching at the Medical College. The experimental results show the excellent performence of the proposed methods. With further research and improvements, the algorithms will be better combined with the platform and be of rich application prospect.