Breast Cancer Diagnosis
Development of a Technology to Create Expert Systems for Diagnosing Oncological Diseases and Demonstrating Their Efficiency Using Breast Cancer as an Example
Tech Area / Field
- BIO-CGM/Cytology, Genetics and Molecular Biology/Biotechnology
- BIO-PAB/Public Health/Biotechnology and Life Sciences
- BIO-PAT/Pathology/Biotechnology and Life Sciences
- INF-IMA/High-Definition Imaging and Displays/Information and Communications
- INF-SOF/Software/Information and Communications
3 Approved without Funding
Belarusian State Medical University, Belarus, Minsk
- Technische Universität München / Institut für Allegemine Pathologie und Pathologische Anatomie, Germany, Munich\nImperial College London / Department of Electrical and Electronic Engineering, UK, London
Project summaryPurpose of the Project – to develop the technology for creating expert systems on the base of a set of quantitative features of cell atypia for diagnosing oncological diseases, to assess the efficiency of the created expert systems using breast cancer as an example.
State of the art in the field of the Project. In the current structure of oncological diseases, breast cancer is the leading pathology among women in the world, mortality to morbidity ratio being 0.35 (IARC, 2004). Analysis of this ratio allowing an assessment of timely diagnosis and quality of treatment points out serious problems in diagnosing breast cancer during the first stage of the disease when the treatment is usually most successful. In Belarus, according to the data from the Research Institute for Oncology, the first stage of the disease with the first established diagnosis of breast cancer is detected only in 19.1% of women [Putyrsky et al., 2005].
One of the main methods for diagnosing oncological diseases is cytological investigation. But cytological diagnosis is problematic due to coincidence of a considerable number of qualitative features of cell atypia in malignant and benign pathology, as well as a subjective factor of assessment of their features. According to different authors, the efficiency of cytological diagnosis ranges from 70 to 80% breast [Hamill et al., 2002; Orell et all, 2005]. Low rate of recognizing breast cancer at the initial stages of the disease shows the need for developing and introducing into practice new and more effective methods of diagnosis. It should be noted that verification of diagnosis at the preoperative stage will make it possible to solve the main question concerning the need for surgical intervention, as well as to determine the strategy, tactics and the extent of operation for oncological patients.
One of the approaches to solving this problem is the transformation of qualitative features of pathological changes in the cells to a quantitative form with the help of computer morphometry. The available literature data testify to the fact that in the studies performed with the help of morphometry, the analysis was conducted as a rule by one sign of cell atypia, particularly, by a quantitative assessment of the size of nuclei, cell clusters, etc. [Baak, 1991]. Moreover, a selective comparative analysis of morphological data performed in these works touched only separate nosologic forms, but not all the main forms of the disease. All this did not allow a significant differentiation of a malignant and benign nature of the disease. At the same time, in the analogy with the method of cytological diagnosis based on a set of qualitative features of cell atypia, it is necessary to create an approach based on a set of quantitative features of cell atypia. The use of a set of quantitative features of cell atypia will allow creating a valuable and objective method of cytological diagnosis. The efficiency of a set of quantitative features of cells in diagnosing thyroid cancer was shown by the project participants earlier [Kirillov et al., 2001, 2003, 2005. 2006, 2007]. Quantitative regularities of pathological changes in the cells of glandular epithelium revealed in the course of the project will make a basis for an expert system for diagnosing breast cancer. While developing the technology for creating expert systems, up-to-date computer methods will be used, allowing automating the process of verification of a diagnosis. All the above – mentioned testifies to the urgency of the planned study.
Impact of the Proposed Project on the Progress in the Field of the Development. The developed technology will allow improving the accuracy, significance and efficiency of cancer diagnosis at the initial stages of the disease and will contribute to further progress in the field of oncology, clinical diagnosis and cytology.
The Project Participants. The project will be implemented by the scientists many of whom participated in the research programs of the military and industrial complex of the former USSR in the field of nuclear and biological weapon, which will allow ensuring their alternative employment. Of 11 scientists and specialists of the 1st and 2nd category, three are Doctors of Science and four – PhDs. The use of their professional knowledge in the field of oncology, morphology and computer technologies for project implementation will make a valuable contribution to solving the problems of practical health care.
Expected results. In the course of implementation of 4 project tasks, the following results are planned to be obtained:
- Experimental model of computer analyzer of images will be created allowing the diagnosis of breast cancer in the automatic mode, and an archive of histological and cytological preparations of breast tissue with malignant and benign pathology will be collected.
- Quantitative regularities of pathological changes in nuclei and aggregates of tumor cells of glandular epithelium in malignant and benign breast diseases will be revealed.
- Expert system for diagnosing breast cancer will be developed and created on the base of a set of quantitative features of tumor cells atypia in glandular epithelium.
- The developed technology will be demonstrated on cytological preparations taken from patients with malignant and benign forms of breast diseases
Scientific significance. As a result of solving the above-stated tasks, quite new technology for diagnosing malignant pathology at the initial stages of the disease will be developed.
Practical significance. Results of project implementation can be used in medical institutions for diagnosing the breast pathology, which will increase the number of patients with the diagnosis of cancer established for the first time during the 1st stage of the disease.
Field of application. The developed technology would be widely used in such fields of medicine as oncology, clinical cytology and diagnostics. Expert systems for diagnosing breast cancer can be used in oncology clinics and dispensaries.
Realization of ISTC aims and tasks. The project meets the ISTC aims and tasks because:
- Project participants from the military and industrial complex will be given an opportunity to direct their efforts to peace activities, particularly, to problems of diagnosing oncological diseases.
- Project participants will be integrated in the international scientific community taking part in the international conferences and workshops related to the problems of oncology, cytology and diagnostics.
- It supports applied studies, as well as the development of technologies for the purpose of peace, especially in the field of oncological diagnosis.
- It contributes to solving national and international problems – diagnosing breast cancer during initial stages of the disease.
- It supports the transition of project participants to market economy by providing an opportunity for creating a software package for expert systems for diagnosing breast cancer, which will be in great demand in medical institutions of different countries studying the problems of clinical diagnosis.
Volume of activities. Duration of the project is 36 months. Project participants include 12 people from the Belarusian State Medical University, 6 of which are “weapon” scientists, 5 – other scientists and specialists and 1 – auxiliary personnel. Total project effort for project implementation makes up 6006 person-days, 3366 (56%) of which are for “weapon” scientists. The project in pided into 4 interrelated tasks. Task 1 includes the creation of a computer analyzer of images and an archive containing 400 cytological and 400 histological preparations. For morphometry investigations, 8 software packages including about 90 main and 200 service functions will be developed. While solving task 2, 30000 nuclei and 20000 aggregates of tumor cells will be estimated using the method of morphometry. Methods for presenting morphometric data will be developed. For this purpose, 4 software packages including 30 main and 70 service functions will be developed. 3 articles and 2 applications for invention will be prepared. Task 3 includes the development of technology for creating expert systems for diagnosing oncological diseases. The system will consist of 2 subsystems. One of them will be developed on the base of a set of quantitative features of nuclei, the other – on a set of quantitative features of aggregates of tumor cells. Formulas for calculating a weighting coefficient for each feature and a diagnostic index will be derived. Instructions for the technology application will be developed with their consequent approval by the Ministry of Health, and 2 articles will be prepared. Task 4 involves the demonstration of the developed technology for diagnosing breast cancer. The efficiency of an expert system will be assessed by 100 samples obtained from different patients and from different clinics. One article will be prepared.
Planned role of foreign collaborators. Foreign collaborators will participate in:
- The development of a work plan of the project;
- The exchange of information;
- Joint workshops.
Technical approach and methodology. Project participants have gained a certain experience in the field of morphology and image processing, which is within the scope of the project. Task 1. Computer analyzer of images will be assembled from purchased parts and supplied with the software for automated morphometry of cytological objects. A software package for image processing and for morphometric assessment of objects will be written in the DELPHI for WINDOWS. Histological and cytological samples taken from patients with ductal and lobular cancer, fibroadenoma and chronic cystic mastitis of the mammary gland will be preparated using standard methods. Two archives will be created on their base. Task 2. The method for providing morphometric data will be developed with the help of the regression analysis. The software for mathematic transformation of the initial base of morphometric data to informative quantitative features will be written in the DELPHI for WINDOWS. With the help of morphometry, nuclei and aggregates of tumor cells of the mammary gland glandular epithelium will be studied. The revealed regularities of pathological changes in nuclei and aggregates will make a basis for the two systems of diagnostic decision criteria aimed at differentiation of malignant and benign nature of the breast disease. Task 3. The technology for creating expert systems for cytological diagnosis of malignant diseases will be developed. For this purpose, standard S-matrix containing boundary values of features for the group of malignant pathology will be created, as well as a research X-matrix containing the values of features of the studied sample. An expert system represents the software, which will allow diagnosing a malignant neoplasm by the value of a diagnostic index by means of comparison of the elements of S- and X-matrixes in the automatic mode. Task 4. The technology developed as a result of solving the first three tasks will be demonstrated while fulfilling this task using breast cancer as an example. Clinical trials will be performed, and the efficiency of expert systems for diagnosing breast cancer will be assessed.
The International Science and Technology Center (ISTC) is an intergovernmental organization connecting scientists from Kazakhstan, Armenia, Tajikistan, Kyrgyzstan, and Georgia with their peers and research organizations in the EU, Japan, Republic of Korea, Norway and the United States.
ISTC facilitates international science projects and assists the global scientific and business community to source and engage with CIS and Georgian institutes that develop or possess an excellence of scientific know-how.