UMD Theses and Dissertations

Permanent URI for this collectionhttp://hdl.handle.net/1903/3

New submissions to the thesis/dissertation collections are added automatically as they are received from the Graduate School. Currently, the Graduate School deposits all theses and dissertations from a given semester after the official graduation date. This means that there may be up to a 4 month delay in the appearance of a given thesis/dissertation in DRUM.

More information is available at Theses and Dissertations at University of Maryland Libraries.

Browse

Filters

2013 - 201912020 - 20241

Settings

Subject: search.filters.subject.Eye tracking

Search Results

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    EFFECTS OF AGE ON CONTEXT BENEFIT FOR UNDERSTANDING COCHLEAR-IMPLANT PROCESSED SPEECH
    (2024) Tinnemore, Anna; Gordon-Salant, Sandra; Goupell, Matthew J; Neuroscience and Cognitive Science; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
    The number of people over 65 years old in the United States is rapidly growing as the generation known as “Baby Boomers” reaches this milestone. Currently, at least 16 million of these older adults struggle to communicate effectively because of disabling hearing loss. An increasing number of older adults with hearing loss are electing to receive a cochlear implant (CI) to partially restore their ability to communicate effectively. CIs provide access to speech information, albeit in a highly degraded form. This degradation can frequently make individual words unclear. While predictive sentence contexts can often be used to resolve individual unclear words, there are many factors that either enhance or diminish the benefit of sentence contexts. This dissertation presents three complementary studies designed to address some of these factors, specifically: (1) the location of the unclear word in the context sentence, (2) how much background noise is present, and (3) individual factors such as age and hearing loss. The first study assessed the effect of context for adult listeners with acoustic hearing when a target word is presented in different levels of background noise at the beginning or end of sentences that vary in predictive context. Both context sentences and target words were spectrally degraded as a simulation of sound processed by a CI. The second study evaluated how listeners with CIs use context under the same conditions of background noise, sentence position, and predictive contexts as the group with acoustic hearing. The third study used eye-tracking methodology to infer information about the real-time processing of degraded speech across ages in a group of people who had acoustic hearing and a group of people who used CIs. Results from these studies indicate that target words at the beginning of the context sentence are more likely to be interpreted to be consistent with the following context sentence than target words at the end of the context sentences. In addition, the age of the listener interacted with some of the other experimental variables to predict phoneme categorization performance and response times in both listener groups. In the study of real-time language processing, there were no significant differences in the gaze trajectories between listeners with CIs and listeners with acoustic hearing. Together, these studies confirm that older listeners can use context in a manner similar to younger listeners, although at a slower speed. These studies expand the field’s knowledge of the importance of an unclear word’s location within a sentence and draw attention to the strategies employed by individual listeners to use context. The results of these experiments provide vital data needed to assess the current usage of context in the aging population with CIs and to develop age-specific auditory rehabilitation efforts for improved communication.
  • Thumbnail Image
    Item
    Reversible jump Hidden Markov Model Analysis of Longitudinal Data with Medical Applications
    (2013) YAN, JIN; Wedel, Michel; Smith, Paul J; Mathematical Statistics; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
    Longitudinal datasets that contain the same variables at multiple time occasions from a given subject are frequently observed in current medical studies. Research has been done to develop method to analyze such data and make meaningful inferences. In this dissertation, we use hidden Markov models (HMM) and a modied reversible jump Markov chain Monte Carlo algorithm to analyze the longitudinal medical data . For an eye tracking data of participants looking at chest X-rays with a potential cancerous nodule, we use the HMM model to nd out what areas on the images attract participants attention more, how their eyes jump among these areas, and which scan pattern is related to an eective detection of the nodule. We estimated the total number of areas of interest (AOIs) on each image, as well as their centers, sizes and orientations. We use pixel luminance as prior information, as nodules are often brighter and luminance may thus aect the AOIs. Dierences in scan patterns between those who found the nodule and those who didn't, are discussed. For a HIV clinical trial data, we use the hidden Markov model to estimate the health states each patient at dierent time points, compare the states with physical phenomena in HIV clinical trials, and predict health development patterns.