Bayesian Statistics Seminar
North Carolina State University
presents
Dr. Sourish Das
SAMSI (Duke University)
"Analysis of 5 Loxin Treatment for Patients with Osteoarthritis in Clinical Trial using Power Filter"
ABSTRACT
5 Loxin is a new and novel treatment for the Osteoarthritis (OA) of knee. In this paper, we present the efficacy study of 5 Loxin treatment for OA and explain the underlying mechanism of a treatment to be successful for treating OA such as 5 Loxin. In a double-blind, randomized, placebo-controlled, clinical trial, the modulation of pro-inflammatory bio-molecules like tumor necrosis factor (TNFa) and interleukin-1b (IL-1b) were evaluated in the OA patients. We present the analysis of association between TNFa and IL-1b with the effect of 5 Loxin treatment for OA. Based on this analysis, we describe the fundamental system of a successful treatment for treating OA, such as 5 Loxin; otherwise any new therapy for OA in future would fail, if they fail to follow the mechanism. We accomplish this by introducing the power filter (PF) for dynamic generalized linear models, which extends the usual Kalman filter (KF) for dynamic linear models. We establish a relationship between the KF and the PF as well. An information processing optimality property of the PF is presented; which shows PF is optimum filter like KF. This optimum property gives PF an edge over the other suboptimal filters, such as particle filter and extended Kalman filter.
Tuesday, September, 30, 2008
4:00 - 5:00 pm
208 Patterson Hall