Lawrence Phillips

Learn More
PURPOSE Decisions leading to drug approval demand careful attention with respect to balancing benefits and risks. The new benefit-risk assessment model provides reassembling of the pieces of decision-making information using computer software to present a coherent overall picture for decision-makers. The aims of the study were to evaluate content validity(More)
The lack of standardized reporting of the magnitude of ischemia on noninvasive imaging contributes to variability in translating the severity of ischemia across stress imaging modalities. We identified the risk of coronary artery disease (CAD) death or myocardial infarction (MI) associated with ≥10% ischemic myocardium on stress nuclear imaging as the risk(More)
Objectives. Determine outcome of the 2005 appropriateness use criteria (AUC) for SPECT in a diverse population of patients and physicians. Background. AUC for SPECT were the first cardiology document to identify 52 clinical indications for imaging, 49 of them for stress SPECT. AUC have been proposed as cornerstone of responsible use of perfusion imaging.(More)
ly, kernel-level rootkits are similar to user-level rootkits. They replace known good state with malicious state. However, the details of kernel-level rootkits are much more complicated than user-level rootkits. Kernel-level rootkits modify running kernel code, which can drastically effect the stability of the system. Previously seen rootkits will modify(More)
BACKGROUND: In order to improve the delivery of health services for chronic medical conditions in our methadone clinic, we added an onsite health screening and brief health counseling to the treatment plans for patients receiving methadone maintenance treatment at the Atlanta Veterans Affairs Medical Center (VAMC). We then conducted a follow up(More)
Statistical learning has been proposed as one of the earliest strategies infants could use to segment words out of their native language because it does not rely on language-specific cues that must be derived from existing knowledge of the words in the language. Statistical word segmentation strategies using Bayesian inference have been shown to be quite(More)
Purely statistical models have accounted for infants’ early ability to segment words out of fluent speech, with Bayesian models performing best (Goldwater et al. 2009). Yet these models often incorporate unlikely assumptions, such as infants having unlimited processing and memory resources and knowing the full inventory of phonemes in their native(More)
Listeners rapidly adapt to many forms of degraded speech. What level of information drives this adaptation, however, remains unresolved. The current study exposed listeners to sinewave-vocoded speech in one of three languages, which manipulated the type of information shared between the training languages (German, Mandarin, or English) and the testing(More)
Significant progress in research has been made in the areas of sex-specific aspects of cardiovascular disease. Despite these advances, coronary artery disease (CAD) is the leading cause of death of women in the Western world. Over the past decade, the focused research on women at risk for ischemic heart disease has helped to clarify our understanding of(More)
Models of language acquisition are typically evaluated against a “gold standard” meant to represent adult linguistic knowledge, such as orthographic words for the task of speech segmentation. Yet adult knowledge is rarely the target knowledge for the stage of acquisition being modeled, making the gold standard an imperfect evaluation metric. To supplement(More)