Scientists at Klick Labs have unveiled a new innovative technique for testing for hypertension using machine learning-enable speech acoustic analysis.
The research findings, published in the journal IEEE Access, hold great potential for a more advanced and accurate early detection of the chronic condition, which the World Health Organization (WHO) terms a “silent killer.”
Although chronic high blood pressure can be life-threatening, early diagnosis can result in timely behavioral intervention, preventing the condition from worsening to becoming fatal.
High blood pressure is widely measured using sphygmomanometry, which requires technical expertise and may not be readily available in remote and underserved areas.
While automatic blood pressure measuring devices can be an alternative, they are also not readily available in such populations, making high blood pressure screening inaccessible in underserved populations.
The Study
The study involved 245 participants, 91 of whom were females. Following the leave-one-subject-out validation, the researchers developed a predictive model for each group based on gender (male and female). The overall results showed an impressive detection rate of up to 77% for males and 86% for females.
These findings could be critical for developing speech-based, non-invasive tests for hypertension and improving access to high blood pressure screening in remote and underserved areas, saving more lives.