Skip to content

Repository files navigation

SpeechDETECT: Speech Analysis Pipeline

Overview

This is the repo of "SpeechDETECT: An Explainable Automated Speech Processing Tool for Early Detection of Cognitive ImpAiRmEnt". This study developed a comprehensive acoustic parameter set tailored to detect speech cues indicative of cognitive impairment. This set evaluates the vocal component across eight domains representing vocal traits—including Frequency Parameters, Cepstral Coefficients and Spectral Features, Voice Quality, Loudness, Intensity, and Speech Signal Complexity—and speech temporal aspects, including Speech Fluency, Rhythmic Structure, and Speech Production Dynamics. We assessed the effectiveness SpeechCARE using the DementiaBank dataset, a recognized benchmark that contains audio recordings from the "Cookie-Theft" picture-description task. Our evaluations, conducted with a machine learning classifier, revealed that Voice-Mark MCI offers significant insights into speech-related cognitive impairments, enhancing early detection strategies for patients at risk of ADRD.

Dataset

We measured the performance of SpeechDETECT using the DementiaBank speech corpus, which includes recordings from 237 subjects who participated in a picture description task. The subjects comprised 122 cognitively impaired and 115 cognitively normal individuals. The dataset was split into training and testing sets with the following characteristics:

Training Data

Attributes Case Group Control Group
Participants 87 79
Gender (F/M) 58 / 29 52 / 27
Age (mean ± std) 69.72 ± 6.80 66.04 ± 6.25
MMSE score (mean ± std) 17.44 ± 5.33 28.99 ± 1.15

Testing Data

Attributes Case Group Control Group
Participants 35 36
Gender (F/M) 21 / 14 23 / 13
Age (mean ± std) 68.51 ± 7.12 66.11 ± 6.53
MMSE score (mean ± std) 18.86 ± 5.80 28.91 ± 1.25

Results

The performance of all machine learning classifiers, following hyperparameter optimization using 5-fold cross-validation, is reported in the following table, based on their evaluation on the test set:

ML classifier F1-score AUC-ROC
Random Forest 63.88 71.58
Extra Trees 60.52 65.39
AdaBoost 68.85 75.55
XGBoost 60.60 62.22
SVM 71.23 77.93
MLP 80.5 79.65

About

No description, website, or topics provided.

Resources

Stars

2 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages