A new study titled Voice for Decision Support in Healthcare Applied to Chronic Obstructive Pulmonary Disease Classification: A Machine Learning Approach explores the potential of using voice technology and machine learning in the early diagnosis of Chronic Obstructive Pulmonary Disease (COPD). The research aims to develop a decision support tool that utilizes voice patterns to classify COPD patients accurately.
Advancements in machine learning techniques offer a promising avenue for harnessing the power of voice technology in healthcare. This study focuses on respiratory diseases such as COPD and investigates the diagnostic potential of machine learning algorithms in analyzing voice patterns linked to the disease. By using the vowel A as a key focus, researchers aim to uncover hidden information that could aid in binary classification of COPD versus no COPD.
The study comprises two main objectives. Firstly, a Systematic Literature Review (SLR) delves into the current state of machine learning algorithms in detecting voice-affecting disorders, identifying existing gaps and proposing directions for future research. Secondly, the study zeroes in on COPD, employing machine learning techniques to analyze voice data with a spotlight on the vowel A. A newly developed Swedish COPD voice classification dataset is expected to enrich the research’s experimental aspects.
The findings from the literature review indicate Support Vector Machine (SVM) classifiers’ prevalence in voice disorder research, with a particular focus on disorders like Parkinson’s Disease and Alzheimer’s Disease. However, gaps persist in terms of underrepresented disorders, limited dataset sizes, and a lack of emphasis on longitudinal studies. In contrast, the research on COPD classification using machine learning techniques presents promising results, shedding light on potential decision support tools for COPD diagnosis.
In conclusion, this study offers a comprehensive overview of machine learning techniques applied to voice-affecting disorders, emphasizing the diagnostic potential of vocal features in healthcare. The research opens up avenues for future technological advancements and underscores the significance of utilizing voice as a digital biomarker for COPD diagnosis through machine learning.
