การพยากรณ์โอกาสรอดชีพของผู้ป่วยโรคมะเร็งต่อมลูกหมาก

Abstract

Prostate cancer there is a high incidence of new cases of males and the mortality rate continues to rise every year. This study examined the survival prognosis of prostate cancer patients. It analyzes survival rates and factors associated with survival of prostate cancer patients and develops a forecast visualization platform. To support the treatment planning of physicians and medical statisticians using data of prostate cancer patients from the Faculty of Medicine Prince of Songkla University. It consists of four factors: Metastasis, Prostate specific antigen (PSA), Age, and Gleason Group for survival analysis. It uses a nonparametric statistical model: Cox model and uses a 5-parametric statistical model: Exponential distribution, Log-logistic distribution, Lognormal distribution, Gamma. distribution and Gompertz distribution. Then evaluate the model's performance. using the Akaike Criteria (AIC). A multivariate analysis of event survival with a Cox proportional hazard model found that the variable that significantly influenced the event-free duration of survival p-value < 0.05 was Metastasis Coef. = 1.842835 (95% CI 1.832- 21.899), PSA Coef. = 0.003220 (95% CI 1.0007- 1.006), Gleason Group Coef. = 0.004026 (95% CI 1.0347- 2.418), Age Coef. = 0.458522 (95% CI 0.9549- 1.056). And the lowest AIC value is the AIC value of the model Gompertz distribution = 206.763. Therefore, it is appropriate to be used to analyze the patient's chances of survival. Then bring the results to convert it into a tabular form. Present information on developed platforms to visualize forecast results by administrators for the user to doctors and medical statisticians’ experiment and consider the results in charts, graphs and descriptive reports. Assessing Usability: Content quality is at a good level (= 4.50/5.00), Design and formatting is at a very good level (= 4.71/5.00), Usefulness is at a good level (= 4.47/5.00) and the Implementation and overall satisfaction of the system. in good level (= 4.50/5.00).