| Abstract [eng] |
Nowadays, medicine plays a particularly important role in our society. The aim is to pass away in peace due to the natural causes rather than a serious or deadly disease. However, data from the Lithuania's National Cancer Institute revealed that the most common oncological disease amongst the men in this country is prostate cancer. Each year, there is an increase in the number of cases of patients with prostate cancer. On the 14th of December 2005, Lithuania's Minister of Health released an order about the early diagnosis of prostate cancer financing program, which stated that the early diagnosis of prostate cancer and prostate-specific antigen detection service will be provided to men aged 50 to 75 years old, and men aged 45 years old if their parents or brothers had a history of prostate cancer. Consequently, due to the rising numbers in the population, there is an increase in the relevance and therefore the effort that is being put into the diagnosis of this disease. This project focuses on the topic of the risk assessment of prostate cancer mortality. The topic is particularly relevant for urologists assessing the survival period of a patient diagnosed with prostate cancer. It would help the doctors choose the most appropriate treatment plan for the patient. The main problem of the study is being expressed in the question: How to quickly and effectively determine the life expectancy in patients with prostate cancer? The objective of the study is the risk assessment of prostate cancer mortality. Considering the subject of the study, the main goal is to develop an automated and statistic-based model to assess the risk of mortality in patients diagnosed with prostate cancer. For achieving this goal, the following research tasks have been set: 1. To analyse researches that were carried out on the topic of the master's final project around the world. 2. To prepare the data set for analysis. 3. To create a research model. 4. To create an interactive system for the estimation of prostate cancer mortality using the developed model and software tools. The first part of the project critically examines the researches that were conducted in relation to this topic. The second part provides methodological information on data preparation, correlation analysis, machine learning, and artificial neural network. In the third part of the final thesis, using the software, based on the created models, the data is analysed, and insights are presented according to the developed models. The created website is being outlined, and the program based on automated and statistical methods is also being described. Data analyses showed that prostate cancer, as a cause of death, remains stable regardless of age group. The created website is easily accessible at: https://indremarijajuskait.wixsite.com/plv-modelis. The program created for this project is found on this website. |
