Individuals with primary progressive aphasia (PPA) present with an insidious onset and gradual loss of word finding, object naming, or word-comprehension skills which profoundly affect their verbal participation in daily activities. The overall goal of this innovative research is to take an initial step toward the creation of adaptive language prostheses that augment lexical access and word use in PPA as skills are lost. The short term objective is to determine whether individuals with mild-to-moderate PPA improve or maintain word finding skills during conversation when provided with a novel intervention tool, namely a mobile technology application called CO-CHAT that automatically presents related vocabulary to them as needed. CO-CHAT is a simulated social media app for research which creates lexical displays synthesized from a user's self-generated photos, comments from social network contacts, the device's metadata, and a curated list of key words generated with Natural Language Processing (NLP) techniques.
Aim 1 addresses development of the simulated social media app with NLP applications.
Aim 2 proposes a research study to determine whether people with PPA can use the CO-CHAT lexical displays to improve or maintain word finding skills in conversation. Two hypotheses will be tested: (1) The number (and percentage) of target words spoken by participants during conversations will increase when the CO-CHAT lexical displays are available~ (2) The number (and percentage) of questions needed by conversation partners to obtain information from participants about daily activities will decrease when the CO-CHAT lexical displays are available. Participants are 10 individuals with mild-to-moderate PPA (agrammatic or semantic variants) recruited from the Oregon Alzheimer's Disease Center. A withdrawal ABAB design with intra-subject and inter-subject replication is proposed. Each participant engages in community- based activities, taking photos and sending them to a simulated social network for comment. By relying on the technology's automatic manipulation of language, photos comments then are analyzed. Related words that are mined from large lexical semantic databases are placed in the lexical displays with the original photo. Participants describe the community activities to familiar partners in 5-minute conversations without technology (baseline phase A) and with CO- CHAT (experimental phase B). Visual analysis of changes across conditions and repeated measures ANOVAs evaluate intervention effects. The proposed research addresses the need to identify effective language compensation strategies to treat individuals experiencing PPA, a relatively new diagnosis for which compensatory treatment paradigms are yet to be developed. Results will support a larger research agenda to further develop adaptive assistive technologies for intervention, and to implement outcomes-based clinical studies that determine the efficacy of a stage-based longitudinal AAC/NLP intervention for patients with PPA in order to maintain vocabulary access, communication functions and social networks with mobile technology over the course of language degeneration.

Public Health Relevance

The population of adults presenting with dementia syndromes and degenerative language disorders is increasing exponentially in the U.S., in the absence of clinical guidelines for effective language intervention. This research will provide evidence to support intervention standards with assistive technologies for persons with primary progressive aphasia, as well as provide scientific data to justify medical insurance reimbursement, and help family members advocate for increases in standard clinical care.

Agency
National Institute of Health (NIH)
Type
Exploratory/Developmental Grants (R21)
Project #
1R21DC014099-01
Application #
8764466
Study Section
Language and Communication Study Section (LCOM)
Program Officer
Cooper, Judith
Project Start
Project End
Budget Start
Budget End
Support Year
1
Fiscal Year
2014
Total Cost
Indirect Cost
Name
Oregon Health and Science University
Department
None
Type
Organized Research Units
DUNS #
City
Portland
State
OR
Country
United States
Zip Code
97239