Brown University Center for Statistical Sciences Seminar
University of Michigan | |
Joint Longitudinal-Survival-Cure Model | |
Abstract: For monitoring patients treated for prostate cancer, Prostate Specific Antigen (PSA) is measured periodically after they receive treatment. Increases in PSA are suggestive of recurrence of the cancer and are used in making decisions about possible new treatments. The data from studies of such patients typically consist of longitudinal PSA measurements, censored event times and baseline covariates. Methods for the combined analysis of both longitudinal and survival data have been developed in recent years, with the main emphasis being on modeling and estimation. We analyze data from a prostate cancer study in which the patients are treated with radiation therapy using a joint model that has been extended by adding a mixture structure to the survival model component of the model. Here we focus on utilizing the model to make individualized prediction of disease progression for censored and alive patients.
In this model each patient is assumed to be either cured by
the treatment or susceptible to clinical recurrence. The
cured fraction is modeled as a logistic function of baseline
covariates, measured before the end of the radiation therapy
period. The longitudinal PSA data is modeled as a non-linear
hierarchical mixed model, with different models for the cured
and susceptible groups. To accommodate the heavy tail manifested
by the data and possible outliers, we use a t-distribution for
the measurement error. The clinical recurrences are modeled
using a time-dependent proportional hazards model for those in
the susceptible group where the time dependent covariates include
both the current value and the slope of post-treatment PSA
profile. Estimates of the parameters in the model are obtained
by the Markov chain Monte Carlo (MCMC) technique. Residuals from
the longitudinal model are plotted for model checking. The model
is used to give individual predictions of both future PSA values
and the predicted probability of recurrence up to four years in
the future. These predictions are compared with observed data
from a validation data set consisting of further follow-up of
the subjects in the study. There is good correspondence between
the predictions and the validation data.
This is joint work with Menggang Yu and Howard Sandler.
**Sponsored by the Charles P. Sisson II Memorial Lectureship
***Co Sponsored by The Marshall Woods Lectureships Foundation of Fine Arts
Cognitive & Linguistic Sciences Colloquium Series
Refreshments will be served before the talk at 3:45 in Room 124-125 |
Abstract: Visual experience consists of more than individual snapshots of the world: we must bind individual views over time into a coherent dynamic experience. Not only must we perceive objects, but we must see them as the *same* objects over time and motion. While a tremendous amount of research has explored static object representations, surprisingly little has focused on the factors which underlie the representation of persisting objects, beyond low-level motion mechanisms. I will describe and demonstrate several new projects from our laboratory which explore three primary aspects of object persistence: (1) Surprising new studies of 'sustained inattentional blindness' and 'motion induced blindness' and 'motion induced blindness', highlighting the extent to which we can completely fail to be consciously aware or salient persisting objects in the first place; (2) Studies of object-specific priming and attentive tracking which begin to reveal the underlying `rules' by which the visual system determines when objects do and do not persist; and (3) Studies of ambiguous motion displays (including work on the perception of causality) which reveal the additional rules which help determine `which went where', in situations involving multiple moving objects.Each of these research strands will involve perceptually salient demonstrations of various types. Most of the experiments and demonstrations will involve adult visual cognition, but these will be related to our other work with human infants and nonhuman primates. Collectively, this work begins to reveal how the mind weaves coherent persisting visual representations out of fragmented snapshots of the world.
Brown University - Computer Science Department Seminar
Host: Professor John Hughes |
Abstract: Film making is undergoing a digital revolution brought on by advances in areas such as computer technology, computational physics and computer graphics. This talk will provide a behind the scenes look at how fully digital films --- such as Pixar's ``Monsters, Inc.'' and ``Finding Nemo''--- are made, with particular emphasis on the role that mathematics plays in the revolution.
Brown Analysis Seminar
Brown University
Joint Materials/Solid Mechanics Seminar
Abstract: `Wet' electrostatic interactions are poorly understood and often exhibit counterintuitive behavior. For example, two like-charged macromolecules can actually attract rather than repel one another and form compact states. From the large body of theoretical work, counterion correlations are thought to play a central role, but there is no consensus for the precise mechanism. We directly measure counterion organization on the surface of actin, a biological polyelectrolyte: Surprisingly, the counterions do not form a lattice that simply follows actin's helical symmetry; rather, they organize into I-D charge density waves parallel to the actin filaments. Moveover, this I-D counterion charge density wave forms a coupled mode with torsional distortions of the oppositely charged polyelectrolyte in order to mediate the attraction. We will also present recent results on applying these ideas to new therapeutic strategies for cystic fibrosis, by examining the electrostatic interactions between biological polyelectrolytes and antibacterial proteins in the airway.
<--- 2004 Index