ASTHMA DETECTION FROM SPEECH SIGNALS

In recent years we find various categories of people suffering with asthma, which is a major cause of death and increased prominence of distress in minor categories of individuals. Numerous studies have also revealed that asthma patients generally have poor drug adherence, and that the intricacy of the regimen may be a contributing factor. The complexity of the regimen and its connection to adherence and asthma outcomes in Ethiopian asthma patients are not, however, known. As a result, this study evaluated how complex medication regimens affected asthmatic patients' medication adherence and asthma control. Vocal cord vibration is a necessary component of speech production. However, asthmatic voice changes will happen because of the inflamed lung airways. Spirometry is a well-known method used to assess a patient's respiratory function and diagnose asthma. Speech data from the subjects included the words "She sells" and the vowel sounds /a:/, /e:/, /:/, /i:/, /o:/, /:/, /u:/, as well as the consonant /s:/. Praat software was used to analyse and create speech parameters from 33 samples. This work specifies the prominence of detection of disease using speech signals. Machine Learning techniques plays a major role in analysing the Categorization of signals using classifiers in Social Spider Optimization in Genetic Algorithm (MSSO-GA) and other feature extraction techniques in Convolution Neural Networks (CNN)