A LOGISTIC REGRESSION MODEL TO PREDICT NODEL MALIGNANCY AMONG CASE WITH PROSTATE CANCER

Abstract

We tackled in this study one of the most outspreading types of cancer among men, that is prostate cancer, which theats a certain age group segment, and mostly it affects those above 65 years. This study was conducted in the Radiation and Isotope hospital – khartoum, with the purpose to predict the spread of the cancer in the lymph nodes among those affected by prostate cancer 60 paients of malignant tumours and we collected the following data from each patient; X-ray results (whether they are positive or negative) , stage (if the disease in its early or advanced stage , blood analysis (PSA) and the age of the patient. We applied this variables on the model which is the subject-matter of this study logistic regression analysis which is consedred one of the best models used for the prediction of these phenomenon. As the dependent variable in this case is the spread of the cancer in the lymph nodes, whether it is positive or negative. If the value of the predicted probability greater than 0.5 this well indicate that the cancer was spread in the lymph nodes . But if the value of the predicted probability less than 0.5 , then this will indicate that the tumour did not spread in the lymph nodes. After the actual application of the data on the model , we found that only the coefficient for PSA (X3) to be significantly different from zero , there are effect of the independent variables ,taken to gether on dependent variable and the logistic regression model is a good fit . From this result we found that the model (logistic regression model) is useful for the study of the relation between the type of lymph nodes (benign or malignant)and the independent variables (age , stage ,PSA and X-ray) i.e. the model is suitable for the study and prediction of this phenomenon. Also, we found that the more the PSA is very high in the blood, then this well serve as a good indication for the spread of the cancer in the lymph nodes and it is one of the most leading factors which could help the doctor to recognize the degree of the spread of the cancer in the lymph nodes.