Projects

Dissertation Projects

Study 1: Physiological Functioning and Later-Life Cognitive Declines

Researchers have explored biological markers to improve Alzheimer's disease identification, but further work is needed to predict cognitive decline and heightened Alzheimer's risk (Alzheimer’s Association, 2019; National Academies, 2017). Further information about these risk factors could allow for modifiable treatment, which may decrease the likelihood of disease development. Past work has linked physiological dysfunctions to differences in cognitive performance in a large cohort of community-residing middle-aged and older adults with cross-sectional data (Karlamangla et al., 2014). This study will examine will assess sub-clinical biological processes in midlife that may contribute to longitudinal cognitive aging and could help identify greater risk of later impairment. 

Study 2: Voice Biomarkers and Later-Life Cognitive Declines

Though researchers have characterized Alzheimer's disease by prosody voice measures, there needs to be more research investigating how these measures can identify higher Alzheimer's risk earlier in life. Further information about these risk factors may facilitate the creation of voice profiles to quickly identify high-risk individuals for intervention. My past work has linked prosody voice measures at occasion 2 to cognitive changes from occasion 1 to 2 in a large cohort of community-residing middle-aged and older adults (Mahon & Lachman, 2022). This study will assess voice changes in midlife previously linked with physiological dysfunction that could help differentiate patterns of longitudinal cognitive aging and risks of steeper decline.

Study 3: Voice Biomarkers and Later-Life Clinical Diagnoses

We already know that prosody voice measures are associated with cognitive impairment in memory and executive functioning (Martínez-Sánchez et al., 2012; Meilán et al., 2012, 2014, 2020). Past research also suggests that voice measures can distinguish healthy cognition from cognitive impairment (Beltrami et al., 2018; Kato et al., 2018; Themistocleous et al., 2020; Thomas et al., 2020; Toth et al., 2017; Xue & Deliyski, 2001). Therefore, using prosody voice measures significant in my previous work (Mahon & Lachman 2022) to predict dementia status could better help identify promising biomarkers for Alzheimer’s risk that can be assessed many years before clinical diagnosis. This study will investigate the relationships between voice prosody biomarkers and dementia status measures at different occasions. This study will be useful in identifying meaningful voice metrics that could facilitate identifying individuals with higher dementia risk.

Automation Projects

Automation Methods

Innovation Project 

Machine learning methods for scoring telephone assessments of reaction time

Issued by Brandeis University (2023) to team:

PI: Dr. Margie Lachman

Supervisor: Liz Mahon

Lead: Sara Motoyama

The Stop and Go Switch Task (SGST), created by Dr. Lachman, uses reaction time in trials requiring attention switching and inhibitory control to evaluate executive function. It is the only test of its kind that is conducted over the phone. This project is a platform that automates this test to make it more adaptable. This research tool will ultimately serve as a catalyst for research in cognitive aging and AD risk, leading to early interventions and improved treatment efficacy.


My Thesis Students

Luna Li: The Role of Anxiety in Voice Prosody and Cognitive Decline in Later Life

Sara Motoyama: Machine Learning Methods for Scoring Telephone Assessments of Reaction Time