Posters
Read research from Cogstate scientists as presented at industry conferences.
A comparison of machine learning based-composite cognitive test scores to track cognitive decline in early stages of Alzheimer’s disease dementia: ADOPIC study
Presented at AAIC 2023
Clinical trials for early dementia require cognitive measures that capture disease progression across multiple domains. Precise cognitive assessment is crucial for tracking early Alzheimer’s disease (AD) dementia. Current cognitive endpoints, like the Preclinical Alzheimer Cognitive Composite (PACC), average standardized change from baseline scores. The use of a machine learning (ML)-based cognitive composite score, computed through principal component analysis (PCA), enhanced the ability to track cognitive decline in individuals with mild cognitive impairment (MCI) progressors compared to PACC, while remaining comparable in cognitively unimpaired (CU) progressors
View PosterAdaptation and Modification of Digital Cognitive Assessments for Smartphone-based and Unsupervised Conduct
Presented at AAIC 2023
Remote and decentralized methods offer advantages in clinical trials, such as reducing burden, improving recruitment, and enabling high-frequency assessment. Smartphone-based cognitive assessment, particularly in BYOD trials, allows for novel designs. However, understanding potential errors from delivery platforms (e.g., smartphones vs. computers) is crucial. Cogstate Brief Battery (CBB) data from young adults in a crossover study (n=60) and a large study (n=35,000) of unsupervised smartphone-based assessments were utilized to gather insights into digital assessments in remote settings.
View PosterAmyloid-β and monocytes in Alzheimer’s disease
Presented at AAIC 2023
This study explores whether human monocyte-derived macrophages can eliminate brain Aβ and migrate to the periphery for antigen presentation, addressing this puzzle.
View PosterCognitive function in older adults with hearing loss; outcomes for treated versus untreated groups at 3-year follow-up.
Presented at AAIC 2023
Cognition remained stable for hearing aid users but declined for non-users at a 3-year follow-up. Hearing aid treatment may delay cognitive decline. Referral for hearing screening and rehabilitation could help minimize cognitive decline in older adults. Hearing loss, affecting 70% of adults aged ≥70, is independently associated with cognitive decline and considered a modifiable risk factor for dementia. Limited evidence exists on the impact of hearing aid use on cognition in older adults beyond 6-12 months, highlighting the need for further objective studies. This longitudinal cohort study compared outcomes of new hearing aid users with those without over a 3-year period
View PosterCross-sectional investigation of synaptic markers Neurogranin and BACE1 in CSF from the AIBL study
Presented at AAIC 2023
View PosterCSF markers YKL40, sTREM2, and a-synuclein enhance the Alzheimer’s disease A/T/N criteria to detect early changes in cognition
Presented at AAIC 2023
Alzheimer’s disease (AD) consists of pre-clinical, prodromal, and clinical phases. CSF biomarkers (A/T/N) aid in characterizing the AD biological state. Assessing eight CSF markers, including sTREM2, α-synuclein, and YKL-40, can predict changes in cognition. These biomarkers may help discern cognitive decline rates in amyloid-positive individuals. Validation of findings is underway in a separate population.
View PosterDiscovering Alzheimer’s disease subtypes with imaging and genetic signatures via multi-view weakly-supervised deep clustering
Presented at AAIC 2023
Alzheimer’s disease and co-pathology exhibit diverse brain atrophy patterns. Genetic variants are associated with heterogeneous imaging patterns in brain diseases. Machine learning methods can analyze neuroanatomical heterogeneity and identify genetically-driven disease subtypes with distinct brain phenotypes. Statistical tests adjusted for APOE e4 were used to identify significant SNPs, providing insights into potential genetic influences on disease characteristics.
View PosterEquating the ADAS-Cog and MMSE to Cogstate Brief Battery scores
Presented at AAIC 2023
The ADAS-Cog and MMSE are standard cognitive outcome assessments (COAs) for Alzheimer’s disease (AD) clinical trials. The Cogstate Brief Battery (CBB) is a digital cognitive test designed for reliable detection of cognitive change. Equating ADAS-Cog and MMSE scores with CBB scores allows validation and translation between traditional and modern COAs. Reliable conversion is observed in clinically better performance, while limited samples in lower performance ranges pose challenges. Comparing traditional and digital assessments aids clinical interpretation and understanding limitations of each assessment.
View PosterFurther Validation of a Self-administered Smartphone-based Verbal Learning and Memory Assessment
Presented at AAIC 2023
The International Shopping List test (ISLT) is a culturally adaptable verbal word list learning assessment. A self-administered smartphone app, ‘List Learning and Memory Assessment Lila TM’, utilizing virtual assistant and natural language processing (NLP), is being validated. Localization of test stimuli and additional data processing options improve accuracy. Reprocessing to address accent, context, and homophones biased to the grocery item has achieved over 97% accuracy.
View PosterIdentification of objective cognitive decline in preclinical Alzheimer’ disease
Presented at AAIC 2023
In Alzheimer’s disease (AD), a slow cognitive decline occurs during a preclinical and prodromal period before dementia onset. Positive AD biomarkers accelerate cognitive decline and progression to dementia. However, distinguishing neurodegenerative disease from normal aging without biomarkers or genetic information is unclear. A measure of objective cognitive decline using flexible mixture models with cognitive composite scores from AIBL shows potential for identifying early cognitive changes and detecting amyloid positivity in cognitively unimpaired participants.
View Poster