Unapređenje procesa detekcije raka dojke primenom računarskog sistema za dijagnostiku integrisanog u medicinski informacioni sistem

Abstract

Breast cancer is the most common malignancy in women. It is often characterized by a lack of early symptoms, which results in late detection of the disease. Detection at advanced stages of the decease implies the treatment is more difficult and uncertain. The treatment of breast cancer significantly depends on the moment of disease detection in relation to the time of the disease onset. The appropriate screening programs have been conducted within the organized preventive examinations and have made significant contributions to the early detection of breast cancer. Unfortunately, a large percentage of the target female population is not covered by these programs for various reasons, and it is necessary to improve the screening process in order to reduce the percentage. Improvement of the existing processes of early breast cancer detection is the theme and the final goal of this doctoral thesis. The success of the entire screening program is conditioned by the quality of each segment of the program, starting from the identification of target population and invitation for the examination, over conducting the examination and referral for further examinations and treatments, up to keeping records and making reports. The improvement of the screening program proposed in this thesis is reflected in the following: more efficient determination of the list of the women who have to undergo preventative examination based on the factors affecting the occurrence of breast cancer, more efficient implementation of screening program in the field, introduction of screening program in thermography as a diagnostic method applied in pre-screening stage, more efficient analysis of mammograms, reporting of the screening program results and continuous follow up of patients. The identification of target population has been based on the age of women. Nevertheless, the fact that a significant percentage of women suffering from this disease are women below 50 years, clearly indicates that some other factors affecting the occurrence of this disease should be taken into consideration during the identification of target groups. This doctoral thesis presents a module of medical information system MEDIS.NET for creating screening list. The system was developed in Laboratory for Medical Informatics at the Faculty of Electronic Engineering Nis, and it is based on the analysis of risk factors affecting the occurrence of breast cancer and enabling easier recognition and classification of women with a high degree risk regardless of age. Breast cancer screening with mammography has been proved very effective in preventing breast cancer death. Nonetheless, due to very complex anatomy of the breast and small differences in density of various breast tissues, the analysis of mammograms is significantly hindered. Thus, the software support to mammographic detection of tumor has become very important. A computer system for tumor diagnosis in mammogram presented in this study includes various methods of image processing, such as methods for classification of mammograms based on texture analysis and methods for segmentation of mammograms, so that the visibility of microcalcifications presenting the earliest breast tumor indicators can be increased. The proposed system provides good results for mammograms with different densities of breast tissue. Early detection of breast cancer is greatly increased by the use of mobile mammograms. Generally, the application of mobile mammograms is often followed by an exceptionally great response of patients. Bearing in mind the necessity for system development in software support to doctors and medical technicians who perform examinations in the field, module of medical information system MEDIS.NET has been developed for supporting the implementation of mobile mammograph that completely overcomes the problems occurring in daily work of radiological services staff in the field. Having considered the fact that the patients covered by the screening program were exposed to harmful X-radiation, the proposed improvement of the screening program includes inclusion of the pre-screening phase in which thermal imaging could be applied. Therefore, the result of pre-screening phases would significantly reduce list of women for screening mammograms. The use of thermal imaging is based on the fact that the vascularization in pre-cancerous tissue is almost always higher compared to the normal breast tissue. This paper describes the computer system for the breast cancer diagnosis based on the analysis of thermal images. The paper shows that the use of classification methods applied for mammograms classification enables successful distinguishing between thermograms with breast cancer and thermograms without cancer, whereas the application of the method for image segmentation in breast thermograms which are recognized as positive enables segregation of almost the exact form of cancer. Introduction of thermal imaging in the screening process is justified by very good experimental results in the form of pre-screening phase.