MODEL MARKOV MULTISTATE PADA PERKEMBANGAN TAHAPAN PENYAKIT DENGAN KESALAHAN KLASIFIKASI

• Main Author: Lestari, Fuji
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/49690
• Institution: Institut Teknologi Bandung

Many types of chronic diseases develop gradually. Therefore, the diagnosis of each stage of the disease is crucial in order to determine the next steps need to be taken in dealing with the disease's condition. However, a classification error on the progression of a disease may occur. This thesis discusses an analysis of a classification error of a disease using a Multistate model based on the Markov process, namely the Hidden Markov model. The Hidden Markov model could estimate the transition rate and the transition probability of the progression of a disease by taking into account the classification error. This thesis also discusses the effect of covariates on the transition rate using the Cox Proportional Hazard model. An example of the application of the Hidden Markov model is given in analyzing the measurement error of the diameter in an Abdominal Aortic Aneurysm. The progression of the disease could be seen from its severity, which is indicated by the changes in the diameter of the aorta. To determine the condition of the widening of the abdominal aorta, an ultrasound test is performed. However, a measurement error could occur in the result of the ultrasound test. In this thesis, an example of the application of the Hidden Markov model is given in analyzing the misclassification of the progression of the Abdominal Aortic Aneurysm disease.