Amyloid Prediction in Early Stage Alzheimer's Disease Through Speech Phenotyping

Novoic Limited67 enrolled

Overview

The primary objective of the study is to evaluate whether a set of algorithms analysing acoustic and linguistic patterns of speech can detect amyloid-specific cognitive impairment in early stage Alzheimer's disease, as measured by the AUC of the receiver operating characteristic (ROC) curve of the binary classifier distinguishing between amyloid positive (Arms 1 and 3) and amyloid negative (Arms 2 and 4) Arms. Secondary objectives include (1) evaluating whether similar algorithms can detect amyloid-specific cognitive impairment in the cognitively normal (CN) and MCI Arms respectively, as measured on binary classifier performance; (2) whether they can detect MCI, as measured on binary classifier performance (AUC, sensitivity, specificity, Cohen's kappa), and the agreement between the PACC5 composite and the corresponding regression model predicting it in all Arms pooled (Wilcoxon signed-rank test, CIA); (3) evaluating variables that can impact performance of such algorithms of covariates from the speaker (age, gender, education level) and environment (measures of acoustic quality).

Study Type

OBSERVATIONAL

Enrollment

Conditions

Alzheimer Disease Preclinical Alzheimer's Disease Prodromal Alzheimer's Disease Alzheimer's Disease (Incl Subtypes)Mild Cognitive Impairment

Eligibility

Sex: ALLMin age: 50 YearsMax age: 85 YearsHealthy volunteers:

Medical Language ↔ Plain English

Inclusion Criteria: * Amyloid status must be known, based on an amyloid PET scan or CSF amyloid test, no older than 30 months at the time of consent for Arm 2 and Arm 4 participants (amyloid negative Arms). * Amyloid status must be known, based on an amyloid PET scan or CSF amyloid test, no older than 60 months at the time of consent for Arm 1 and Arm 3 (amyloid positive Arms). * Subjects must be aged 50-85 (inclusive). * Subjects must have MMSE scores of 23-30 (inclusive) based on a test not older than 1 month at the time of the visit. * Date of diagnosis (if applicable) maximum of five years prior to consent. * Subjects' first language must be English. * Willing to participate in a study investigating speech and dementia. * Availability of a person ('caregiver') who in the investigator's judgment has frequent and sufficient in-person contact with the participant, and is able to provide accurate information regarding the participant's cognitive and functional abilities. This is most likely met when living with a caregiver. * Able to provide valid informed consent. * Able to use, or has a caregiver who is able to use a smartphone device. * Has access to a smartphone device running an operation system of Android 6 or above; or iOS 10 or above. If taking part in the study through virtual visits, the following inclusion criteria also applies: * Able to use, or has a caregiver who is able to use a personal computer, notebook or tablet. * Has access to a personal computing device of that is: * Running an operating system of macOS X with macOS 10.9 or later; or Windows 7 or above; or Ubuntu 12.04 or higher; or * Have access to one of following internet browser software Internet Explorer version 11 or above; or Microsoft Edge version 12 or above; or Firefox version 27 or above; or Google Chrome version 30 or above; or Safari version 7 or above; capable of audio and video recording; and able to connect to the internet. Exclusion Criteria: * Clinically significant unstable psychiatric illness in 6 months. * Diagnosis of General Anxiety Disorder. * Current, or history within the past 2 years of major depressive disorder diagnosis (according to DSM-5 criteria); or psychiatric symptoms that, in the opinion of the investigator, could interfere with study procedures. * History or presence of stroke within the past 2 years. * Documented history of transient ischemic attack or unexplained loss of consciousness within the last 12 months. * The participant is using drugs to treat symptoms related to AD, and the doses of these drugs were not stable for at least 8 weeks prior to consent.

Outcomes

Primary Outcomes

Area under the curve (AUC) of the receiver operating characteristic (ROC) curve of the binary classifier distinguishing between amyloid positive (Arms 1 and 3) and amyloid negative (Arms 2 and 4) Arms using speech recordings as input.

Time frame: baseline

Secondary Outcomes

The sensitivity of the binary classifier distinguishing between amyloid positive (Arms 1 and 3) and amyloid negative (Arms 2 and 4) Arms.

Time frame: baseline

The specificity of the binary classifier distinguishing between amyloid positive (Arms 1 and 3) and amyloid negative (Arms 2 and 4) Arms.

Time frame: baseline

The Cohen's kappa of the binary classifier distinguishing between amyloid positive (Arms 1 and 3) and amyloid negative (Arms 2 and 4) Arms.

Time frame: baseline

The sensitivity of the binary classifier distinguishing between amyloid positive cognitively normal (CN) (Arm 3) and amyloid negative cognitively normal (CN) (Arm 4) Arms.

Time frame: baseline

The specificity of the binary classifier distinguishing between amyloid positive cognitively normal (CN) (Arm 3) and amyloid negative cognitively normal (CN) (Arm 4) Arms.

Time frame: baseline

The Cohen's kappa of the binary classifier distinguishing between amyloid positive cognitively normal (CN) (Arm 3) and amyloid negative cognitively normal (CN) (Arm 4) Arms.

Time frame: baseline

The AUC of the binary classifier distinguishing between amyloid positive cognitively normal (CN) (Arm 3) and amyloid negative cognitively normal (CN) (Arm 4) Arms.

Time frame: baseline

The sensitivity of the binary classifier distinguishing between amyloid positive MCI (Arm 1) and amyloid negative MCI (Arm 2) Arms.

Time frame: baseline

The specificity of the binary classifier distinguishing between amyloid positive MCI (Arm 1) and amyloid negative MCI (Arm 2) Arms.

Time frame: baseline

The Cohen's kappa of the binary classifier distinguishing between amyloid positive MCI (Arm 1) and amyloid negative MCI (Arm 2) Arms.

Time frame: baseline

The AUC of the binary classifier distinguishing between amyloid positive MCI (Arm 1) and amyloid negative MCI (Arm 2) Arms.

Time frame: baseline

The sensitivity of the binary classifier distinguishing between the MCI (Arms 1 and 2) and the CN (Arms 3 and 4) Arms.

Time frame: baseline

The specificity of the binary classifier distinguishing between the MCI (Arms 1 and 2) and the CN (Arms 3 and 4) Arms.

Time frame: baseline

The Cohen's kappa of the binary classifier distinguishing between the MCI (Arms 1 and 2) and the CN (Arms 3 and 4) Arms.

Time frame: baseline

The AUC of the binary classifier distinguishing between the MCI (Arms 1 and 2) and the CN (Arms 3 and 4) Arms.

Time frame: baseline

The agreement between the PACC5 composite and the corresponding regression model predicting it in all four Arms, as measured by the coefficient of individual agreement (CIA).

Time frame: baseline

For each classifier/regressor in outcome 1-16, the correlation between the AUC/CIA and each age group, gender and speech-to-reverberation modulation energy ratio group, as measured by the Kendall rank correlation coefficient.

Time frame: baseline

Amyloid Prediction in Early Stage Alzheimer's Disease Through Speech Phenotyping

Overview

Conditions

Eligibility

Locations (1)

Outcomes

Primary Outcomes

Secondary Outcomes